Bearhawk Aircraft Bearhawk Tailwheels LLC Eric Newton's Builder Manuals Bearhawk Plans Bearhawk Store

Announcement

Collapse
No announcement yet.

Airfoils, Ordinates, Fat Wings, Harry Riblett and other musings…

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    I have obtained some additional information since my last post that takes a bit of air from under my wings and leaves me a little less excited about the path that I was planning to take in forming the ribs.

    It turns out that the dimensional discrepancies in Dwg 4 are not printing or drafting errors but rather an intentional modification to the wing profile. I have learned that the modification, in some form, also exist on the mylar of the Patrol, 4-Place and possibly the Model 5, as they share the same profile and chord dimension. I don’t have the mylar for any of those aircraft so I can’t verify what I learned from other sources. I have calculated and plotted the GA30-413.5 airfoil with a 65.125 inch chord should anyone want to do the comparison. Link below. So, the answer to the obvious question seems to be completely unknown to anyone except Bob, and he appears to be deflecting or obfuscating, based upon my research and the answer obtained recently by Mark, on my behalf.

    If the modification was done to improve performance or flight characteristics, then why did previous members not receive that information as an answer to their questions years ago? I’ve also learned that some aircraft designers, including Steve Wittman, would sometimes attempt to smooth-out the concave portion of a high camber wing in order to make rib forming a bit easier. If that was Bob’s reason for the modification then why not tell us? I can also speculate on a couple more possibilities but sometimes speculations can be interpreted as accusations and I certainly don’t mean to infer anything negative.

    If the modification was made in order to make rib forming easier for plans builders, then what about the factory ribs? They were formed in the -O condition, thus much easier to form, so do they have the correct 613.5/413.5 profile or are the ribs formed to their respective mylar profile? The LSA is marketed as having the Riblett GA30-613.5 rib but that is only partially correct as it only applies to the plans set drawings. If the factory ribs are formed to the Dwg 4 dimensions, that marketing statement would not be correct as the Dwg 4 ordinates are not obtainable by calculations using Harry’s GA30-015 basic thickness form data with the GA-6 camber data. It might, then, be more correct to refer to the wing rib profile as a Bob Barrows modified version of the GA30-613.5 instead.

    Having purchased our chosen aircraft plans set, does that cost not include the information that we might want or need that would further our knowledge and understanding of that aircraft and, specifically, what modifications were made to the airfoil that technically changed it from a true 613.5?

    I might be the only current forum member that would like an answer to these questions but, based upon my research, there have been many more over the years. There are, likely, only a handful of people on this forum that could ask these questions to Bob with the persistency possibly needed to get an answer without creating animosity. And none of those are me…

    Best regards to all,
    Mitch



    P.S. Until a few days ago, I was assuming that I would be forming my ribs to the true 613.5 profile. Maybe still. Although I planned to have the master template CNC cut, I still wanted to have a mylar drawing to use as a reference at times and to maybe hang up on the shop wall. Duncan-Parnell was my source for topographical maps back in the late 70’s and I knew that they were still in business and now had a location in Concord. Still a 45 minute drive but a nicer drive than to Charlotte. They only had 24” mylar so I put two different profiles onto one 65” length. It was $19.00 plus tax. Not too bad, I don’t think. I have included a link to the PDF that I had printed should anyone like to have it for any reason. It might make a good conversation piece for your workshop. It cost about $6.00 on standard paper.


    Riblett GA30-413.5 Airfoil Profile - http://rf-tech.us/lsa/Riblett_GA30-4...art_Points.pdf

    Table of Chord Stations and Ordinates for the 413.5 Airfoil - http://rf-tech.us/lsa/Riblett_GA30-4..._Ordinates.txt

    Riblett GA30-613.5 Airfoil Profile and Chord Station Dimensions - http://rf-tech.us/lsa/Riblett_GA30-6...l_Profiles.pdf
    Last edited by NCLSA183; 09-13-2021, 08:40 PM.

    Comment


    • NCLSA183
      NCLSA183 commented
      Editing a comment
      Hello Frank. Good to hear from you. I hope your build is progressing nicely. I do hope to get started banging ribs before the end of the year. Still getting my ducks lined up and just stirring the pot while I'm waiting. Have you ever cooked a big cast iron pot of chili beans over an open fire? You got to keep stirring or those suckers will stick to the bottom and create a big mess!! I promise you, it is not my intention to stir the airfoil pot and, more seriously, not to offend anyone. You may remember from my original post, I learned about the Dwg 4 issue, or whatever we may call it, about 4 years ago and I can not find any resolution that is documented. I'm pretty darn sure that there is a reason behind it but I can't find an answer. I don't mind calling Bob but based on his response to the question over the years, I'm relatively certain that he wouldn't be forthcoming. I did send an email 2-3 weeks ago, no response, but I will not rule out that phone call, eventually. Maybe even a physical visit. He isn't' too many hours away and seems like a nice guy to chew-the-fat with.

      Frank, the Riblett airfoil profile GA30-613.5 is mentioned many times in print and videos and also Page 26 of the LSA Builder Manual that came with the plans. I'm confident that the aircraft is and would be a great performer with the 613.5 profile or the Dwg 4 profile. I also wonder if the 613.5 was flight tested before making the Dwg 4 change. Dang, another question waiting for an answer. The LSA plans were drawn up and the loads were calculated using the true 613.5 profile but Dwg 4 departs from that profile and there is what has stimulated my profound curiosity. I suppose if Bob were to tell us that is was none of our business as to why he modified the profile, then I would likely stop stirring but I would still be pretty-darn-curious.

      Take care and keep banging!!

      Mitch

    • Frank
      Frank commented
      Editing a comment
      Ha! Shows how much I know! My excuse, just getting to page 10: "Once the ribs are done you can start to work on the spars." When I get back from short vacation I'll download your attachments for my files.

    • NCLSA183
      NCLSA183 commented
      Editing a comment
      Hope you have a wonderful vacation, Frank.

      I just want to be clear to all members that the profile drawings that I have linked are not meant as a replacement for the Dwg 4 mylar sheet. They may be nothing more than a worthless souvenir from my recent Bearhawk safari.

      I do have a set of ordinates I arrived at mathematically that will duplicate the profile of Dwg 4. I created it as an educational project, and for my own use, but I would like to share with other LSA builders who, like you and me, might prefer to preserve their mylar and just use a paper copy to create their master template. While I see no legal reason that I cannot share this data with licensed plans owners, there is an ethical issue that I can't overlook.

      Mitch

  • #17
    I've been inactive or lurking a bit (remodeling the house instead of building airplane parts after a flood in the kitchen), but found this thread intriguing. I too found that my estimate of the Riblett 13.5% thick airfoil from the description of how they were built from his book vs. the as-built to be different. I took a very low-tech method of measuring the Mylar drawing by hand and comparing it to my estimates based on Mr. Riblett's descriptions of how he modified the NACA mean line and thickness forms (I never did use any of the calculators you can find online, but I was able to replicate his published thickness forms, so I think I got it right. I'm attaching an image of my analysis - showing the theoretical Riblett 13.5% thickness form, mean line, and overall airfoil ordinates vs. my measured values. You'll see that they're different.

    Now... at the risk of a little hubris, I am an aerospace engineer, I work in the field of aircraft design and performance analysis, I've taken airfoil design and analysis classes, and I use aerodynamics every day as part of my day job (okay, when I'm not answering e-mails, writing reports or PowerPoint charts, preparing budgets... I guess I do less analysis lately than I wish I could do these days). I've analyzed the Riblett and Bearhawk airfoils with better computational tools than Riblett used, and when I get the time, hope to run full Navier-Stokes simulations on the airfoils. I even have visions of a slotted flap design that uses the same hinge point or a drooped hinge point, but that takes time that I don't currently have for what is likely a dubious return on investment. The best thing the Bearhawk has going for it is a solid history of good performance. I'll take the factory form of the Bearhawk airfoil any day over the theoretical form of the 13.5% airfoil. I do recall chatting with Bob when I picked up my plans, and we spoke a bit about the airfoil selections and aerodynamics when he heard of my day job. I believe he mentioned that Mr. Riblett had sent him the airfoil ordinates, and I believe had even suggested some modifications (I don't know if it was Bob or Riblett that suggested the mods, but it does appear they were made).

    At the end of the day, I'll trust the scoreboard - Bob's design clearly does well, even if it doesn't match the theoretical ordinates implied by the Riblett text. I don't think it matters that it doesn't match the Riblett ordinate descriptions/formulas - it clearly works. Besides, I suspect that Mr. Riblett and I would not see eye-to-eye on everything, and I find some of his assertions a bit dubious, but I do agree with him more than I disagree. Sadly, I'll never get the chance to make my case to him on this earthly plane.

    airfoil_form_comparison.png
    Attached Files
    Last edited by nborer; 09-20-2021, 11:38 PM.
    4-Place Model 'B' Serial 1529B (with many years to go...)

    Comment


    • nborer
      nborer commented
      Editing a comment
      I believe the 4-place "A" model uses a NACA 4412. The modified Riblett 13.5% was introduced on the Patrol, and then used for the 4-place "B" model. I don't know what is used in the LSA, Companion, or Five, but I think it's reasonable to believe that they all use the same airfoil as the Patrol and the B.

    • NCLSA183
      NCLSA183 commented
      Editing a comment
      Hello nborer. I’m sorry to hear of your recent misfortune with flooding but I’m glad that you were intrigued enough to offer your analysis and/or opinions. Your “low-tech” method would hardly seem as such to many of us. I did enjoy the time that I spent in deciphering the graph of your comparisons. It’s a bit above my paygrade, though.

      It seems that everyone, including myself, agrees that the Beakhawk aircraft have great flying qualities, although I have never flown in any of them. You didn’t state your experience, either way. But, that was never the intended point of this conversation and would surely be very subjective, besides. I certainly can’t say that I would accept the scoreboard when there’s only one team on the court. I will attempt to explain…

      Let me start by respectably disagreeing that the 30-613.5 airfoil is “theoretical” as it has been flight tested on several different aircraft, since first published by Mr. Riblett, including the very popular Highlander STOL. While not being an apple-to-apple comparison with the LSA, all these aircraft have great flying qualities.

      This leads me to restate my intended point of this conversation, with modifications and bullet points:

      * The Bearhawk LSA wing was designed, drawn and load calculations performed using the true 30-613.5 profile. This can be easily verified by using the printed ordinate data from Mr. Ribletts book or by calculating the ordinate data, oneself, using his 30-015 basic symmetrical form, GA-6 camber data, scaling to a 60 inch chord and comparing these data dimensions to LSA Dwg 3.

      * The Bearhawk LSA wing was never built and flight tested with the as-designed, true 30-613.5 profile. Therefore, no flight data was available to be recorded, compared or shown on the metaphorical scoreboard. No cheerleaders, either...

      * Without building and flight testing the as-designed, 30-613.5 wing, a modification was made to the airfoil profile for reasons unknown except to Bob or Harry, as far as I know. Without data to compare, no one can really know if the result of the modifications were as intended. It’s possible that the modified wing could be less desirable than the 30-613.5 wing would have been. Especially, considering how well the 30-613.5 has performed on other aircraft.

      * While we don’t know if the Dwg 4 wing is any better than the 30-613.5 wing, we do know that it is thicker, heavier and, most likely, has more aerodynamic drag. The Dwg 4 wing probably has slightly less camber that would, in my opinion, result in a marginal speed increase but so might the speed be higher with the lighter, thinner and less aero drag 30-613.5 wing. The higher camber 30-613.5 should have a slower stall, steeper climb than the Dwg 4 wing. Again, my opinion from my research.

      * Since we don’t know the intended purpose for the modification and we don’t have data to compare it to the true 30-613.5 wing then we, I, can’t make an informed decision as to which airfoil I want to build. I’m relatively sure that Bob never meant for us to have a choice but, yet, has unwittingly given us one by not being forthcoming with the information that many of us have asked for, over the years. While he isn't wrong, his recent response to the question was, effectively, "use drawing 4, it flies just fine, stop asking questions". I took no offense but was a bit disappointed.

      I, for one, would love to have someone with your expertise and experience to analyze the two different profiles of the LSA. From my understanding, it may also apply to the airfoil used on the Patrol, BH4B and BH5 due to a similar modification of the true 30-413.5 profile. Not that you would need my assistance, but I would be glad to furnish you with the ordinates for both the 413.5 and 613.5 profiles in their respective chord lengths, as well as, the ordinates for the LSA Dwg 4 mylar profile. Or any other info than I may be of help with. The Navier-Stokes simulation now has me very intrigued.

      Thanks again for your input!
      Mitch

    • NCLSA183
      NCLSA183 commented
      Editing a comment
      Frank,

      The Patrol, BH4B and BH5 use the GA30-413.5 airfoil, modified I'm told, with a 65.125 inch chord. The LSA uses the GA30-613.5, modified, airfoil with a 60 inch chord. The Riblett airfoils are scalable to different thickness distributions and chord length within the different series. I have created a spreadsheet that makes those calculation's very simple by only inputting the thickness distribution percentage, i.e.13.5%, and a chord length, i.e. 60 inch. I have been meaning to share a link to the file but I wanted to create some basic documentation so that it would be usable by most anyone regardless of their knowledge of Excel. I'll try to get that completed and posted ASAP. Hopefully, nborer might find some time to help with the analysis that you speak of. We have to keep stirring...

      Mitch

  • #18
    Mitch:

    Thank you for the consideration. I'll pose a few responses to your assertions that I hope will motivate some additional thoughts and perhaps help you in your quest for the best way to move forward.

    You mention that the 30-613.5 airfoil is not theoretical, "as it has been flight tested on several different aircraft, since first published by Mr. Riblett, including the very popular Highlander STOL." I agree that the airfoil characteristics have been "demonstrated" in flight. However, I have yet to see actual *test* data for that airfoil, or, frankly, for any of Mr. Riblett's airfoils. When I work on a project and have a requirement with a stated verification method of "test," I am expected to conduct a highly controlled experiment to determine if the requirement is met. A verification method of "demonstration" indicates that far more subjectivity is involved. I've seen dozens of "demonstrations" of the effectiveness of Riblett airfoils - pilot testimonials, reports, even quotations of numbers associated with changes in stall speed, stall characteristics, cruise performance, etc. However, I have yet to see a highly controlled set of experimental data associated with any Riblett airfoil - can those providing those testimonials show me a data sheet indicating the density altitude, weight, CG, and all other variables associated with the aircraft prior to the modification are the same? Can they provide how they calibrated their equipment? Did they use best practices for evaluated handling qualities or task completion, like Cooper-Harper ratings? Please, if you are aware of any, please send my way - this is not snark, but I have not seen anything about these airfoils tested in a controlled environment. A "test" should try to reduce all uncertainty - discuss ways in which the environment is controlled and/or measured, the manner in which the test article is controlled (including the tolerances and verifications associated with those tolerances), the approach for test equipment calibration, and even how the objectives of the experiment are designed. If not, it is a "demonstration" - something that can be very powerful (and is used very often in engineering!), but something with far less rigor.

    I know that sounds snooty, and that is not my intent. I have read Riblett's book - in whole or in part three times - and I largely *agree* with his overall assertions regarding airfoil design practices. Unfortunately, he disparages a lot of good research conducted by honest folks over many decades, and I'm happy to offer some alternate explanations as to why things ended up the way they did that are less... conspiratorial. I personally know two of the individuals that have worked on the airfoils or GA projects that he ends up maligning in his book, and I think his narrative does not provide the appropriate context. That said, it's his book, not mine, and he's free to his thoughts and opinions. Beyond that, I won't go into the weeds here and I have no wish to argue with someone who is not around to defend himself.

    Going back to *test* data... from my (admittedly limited) searches, and from reading Mr. Riblett's book (I have the sixth edition), all of the airfoil profiles were generated per a (sound) theory, and the performance was estimated for these airfoils with a dated, but reliable, analysis code. However, I have not seen reliable, repeatable, calibrated *test* data associated with any of these airfoils. Again, I'm happy to be shown otherwise, but all I've seen thus far are very positive results from demonstrations, not from tests. In the absence of this data, any discussion that the 30-613.5 airfoil is somehow tested and vetted is moot - it has been demonstrated (very!) favorably, but I have not seen calibrated data that validates performance claims - or, even, a thorough description of those detailed performance claims. I'm not referring to "gentle stall," I'm referring to "stall at an angle of attack of 19 degrees with a pitching moment about the quarter chord of -0.026 when corrected to a standard atmosphere at a Reynolds number of 3,000,000."

    You state that the "the Bearhawk LSA wing was designed, drawn and load calculations performed using the true 30-613.5 profile" and claim that this can be verified by inspecting LSA drawing 3. I don't know LSA drawing 3, but if it's similar to my model B plans, drawing 3 shows various wing ribs (nose, center, etc.). I would not consider those to be a definitive source on rib geometry - I would consider the formblock master, the Mylar sheet that is drawing 4 in the model B plans, as the source for the rib shapes. That is the intent of the Mylar sheet - in fact, it is on Mylar because, as a medium, it is subject to less deformation than the paper drawings. My first internship was as a drafting/CAD monkey, and I had to edit the official drawings in our computers (and some older hand-drawn line drawings) to what the guys on the shop floor had marked up and were actually using. I learned quickly that some drawings and dimensions "controlled" others, beyond the standard that we learned in class regarding controlling dimensions on an individual drawing. I can offer you no objective proof that the Mylar drawing is the controlling drawing, other than it makes the most sense and fits anecdotally with what I have read on this site.

    Regarding the load calculations, Riblett never provides loads in his book for the 30-613.5 airfoil. He provides a scaled picture of its shape and an estimate from his computer code for minimum drag, maximum lift, and pitching moment about the quarter chord at some selected Reynolds numbers in Table I and Figure III-19 on page 71 of my edition, but the only way to estimate lift, drag, and moment coefficient per the data he provides for the GA30-613.5 airfoil can come from interpolating the plots on page 77. By my own estimates from XFOIL, the small changes in lift and drag from the Bearhawk Model B drawing 4 (Mylar) profile and the published Riblett method for defining that airfoil geometry for the GA30-413.5 are so small as to not warrant further consideration. The only larger difference I have is in moment coefficient, and I suspect that lies in the discretization of the Riblett book ordinate definition vs. the finer discretization (and associated numerical smoothing necessary) to eliminate numerical roughness that I applied to my personal measurements but did not apply to the Riblett book measurements (the variations in moment coefficient otherwise defy conventional logic). I won't get into such nuances here. We should also recall that the 4-place wing structure was designed for the thinner, more highly loaded NACA 4412 profile, so that would imply to me that applying the same skin and spar cap thicknesses would result in higher structural margins, since the Model B airfoil is thicker. I'm attaching a PDF of my XFOIL estimates for the "book" Riblett GA30-413.5 airfoil, my Kulfan-smoothed estimates from the Model B drawing 4 (Mylar) airfoil, as well a stock NACA 4412 (similar to Model A?) airfoil to this message. I'm also providing a revised estimate of the camber line and forms for the book vs. smoothed drawing 4 airfoils - my previous post was a bit confusing and based on data I hadn't looked at in 18 months. I've removed the x/c and y/c axis values and made sure the scaling in x and y isn't square so I don't run afoul of any copyrights.

    Bringing it home... I don't know how many Bearhawk LSAs have been built and flown, either scratchbuilt or from the factory. But I'll bet you a beer that the ones that have been built and flown have used the Mylar drawing as the formblock, and not the published approach to the Riblett 30-613.5 ordinate generation (let's recall that Riblett told us how to generate thicknesses in between the 12%, 15%, and 18% forms in his book, but he did not explicitly publish the ordinates for anything in between - though I suppose you could scale up the picture he provides in Figure III-19). By that standard, they fit the "demonstration" criteria that they fly well - just as much as other "testimonials" that the published approach to generating the airfoil ordinates produces good flying characteristics. I'll bet you another beer that whatever ordinates show up in the Mylar drawing for the LSA will perform within 5% of the lift and drag estimates that one would reasonably expect using modern-day analysis techniques on both sets of ordinates.

    If you wish to send me your measurements of the LSA, I can compare those to my ordinates for the GA 30-613.5 airfoil. It would probably also help to have your ordinates based on Riblett's book, to make sure we're both working from the same sheet of music. I can run a similar comparison as I've done here for the GA 30-413.5 airfoil and the Bearhawk Model B drawing 4 airfoil. PM me if it's more straightforward or if you're concerned with copyrights.

    Nick
    Attached Files
    Last edited by nborer; 10-01-2021, 10:56 AM.
    4-Place Model 'B' Serial 1529B (with many years to go...)

    Comment


    • NCLSA183
      NCLSA183 commented
      Editing a comment
      Nick… Dang!… Where do I start??... I have “information overload”!! It would appear that you must have missed a previous post where I placed the disclaimer “I am not an engineer”!! LOL!! Unfortunately for you, I do still possess that “yearn to learn” mentality and, as time permits, I’ll be spending some of that time with Mr. Google. The “unfortunate” part of that statement will likely be the volume of questions that I’ll have for you at a later time. LOL, again!

      We are working from the same sheet of music, in regard to Mr. Ribletts’ book, but your apparent interpretation of some of my statements would seem to be a page or two off from what I had intended or hoped. I have never considered myself to be an articulate communicator and that may have contributed to a possible misunderstanding or, at the least, suggest some clarification.

      When I disagreed that the 30-613.5 airfoil was “theoretical” I was only referring to the fact that it was no longer a theory in someone’s head or a sketch on a napkin but rather had reached the point of practical application on several highly regarded aircraft. I believe that to be a valid and adequate reason to un-characterize the airfoil as theoretical.

      You may have taken some issue with my use of the term “fight tested”. In hindsight, and by trying to look at it from your perspective, I can understand if you did. I only meant to convey my thinking in that the 30-613.5 airfoil was flight tested in the same manner as all the Bearhawk airfoils and most other homebuilt or low production aircraft, i.e., pilot testimonials, and that the subjective data from the 30-613.5 testing were equally as compelling as the Bearhawk airfoil data. That is why I said that I couldn’t accept a scoreboard with only one team competing. From an engineering standpoint, I completely agree that, demonstrated characteristics, would be a better term to use and that a controlled flight-testing regime, as you presented, would certainly be ideal. But for now, I suppose our best option is to hope for comprehensive, unbiased flight test reports from seasoned and respected test pilots.

      Yes, Mr. Riblett could have been much more respectful of the oftentimes invisible but yet very talented and hardworking individual’s that make up a large government entity such as NACA/NASA. I make no excuses for him on that point. Throughout this topic of discussion, I have attempted to not misled anyone into thinking that I have expertise in any aeronautical discipline or that I am qualified to offer any arguments/counterpoints to Mr. Ribletts design theory, methods or performance analysis. I would like to learn more about this interesting topic but I don’t currently have the time to invest. If I unknowing implied that I had seen or heard of any reliable, repeatable, calibrated test data of the 30-613.5 airfoil, it was completely unintentional. I, too, wish that it existed. But it has been tested/vetted to the generally-same extent as the modified airfoils of the Bearhawk line and that was my only available method for comparison.

      I am guilty of assuming that a contributor to this topic/discussion has read and remembered the, subjectively, important bits. Because of that, I tend to not go into complete detail when I attempt to explain or make points of my findings, opinions or speculations. That is meant as a courtesy to anyone that is sacrificing their time to read my often, wordy musings. But, it can, and likely does, sometimes result in our message not being transmitted clearly and/or received without overriding background interference.

      A couple months back, after having decided that I might be able to get started on my LSA by years-end, I started thinking about this well-known question regarding the airfoil for the LSA and others. I had a thought process already in my head so I pulled out Mr. Ribletts book and started some heavy reading. I didn’t understand a great deal of the data graphs and, due to a very blurry copy of the book, I couldn’t read the numbers and text within many of the graphs so I didn’t spend a lot of time trying. Nick, unless I completely misunderstand your comment about there not being any published ordinates for a thickness distribution that fall within the 12%, 15% and 18% ordinates, my copy, 6th edition, has the 13.5% ordinate table on Page 71 just above Figure III-19 that you reference in your most recent post. I used those published ordinates, scaled to a 60” chord length, to compare with the rib profile drawings of LSA plans sheet Dwg 3 and the Main/Rear/Aileron spar dimensions of Dwg 3 & Dwg 5. I made very careful and repeated measurements of these drawn wing components with a digital caliper. The rib profiles on Dwg 3 include the flanges and match up near perfectly with the published ordinate dimensions from Riblett’s book and at each station point along the chord line, allowing for manual drafting tolerances. The accuracy was so near perfect that I even made a comment in one of my earlier postings that Bob was a virtual Ninja at the drafting table. In addition, the measured thickness of the center rib at the main spar attach point was within a few 1/100’s of the call-out dimensions of the main spar height as shown on Dwg 5. This is what I considered and described as verifiable data to confirm that the LSA wing was indeed designed around the true as-published and scaled GA30-613.5 airfoil profile.

      Regarding the wing load calculations, I could have done a better job in my wording from my last post. I seem to remember that I described it a bit differently at an earlier time but that point is mute. So, after I was convinced that the as-designed wing profile was a verifiable 30-613.5 airfoil, I then postulated that Bob would have also, most likely, used the spar dimensions from Dwg 5 for his wing load calculations. Verifiable only by Bob, I suppose.

      Soon afterward, I used Harry’s guidance and created a spreadsheet that would allow me to calculate, scale and create a DAT file for any combination of thickness percentage and chord length within a series. What range of combinations would be useful, I haven’t a clue. I tested the spreadsheets accuracy with Harry’s procedure of using his basic, symmetrical thickness form, GA30-015, along with the camber profile ordinates GA-4 and GA-6. By using the appropriate inputs for percentage, chord and camber, the spreadsheet will output ordinates that match all published ordinate tables for the chosen airfoil series. Satisfied with this, I then plotted and printed the airfoil profile for the GA30-613.5 with a 60-inch chord. Overlaying this printout with the Dwg 4 mylar shows a perfect match in all areas except for the lower most perimeter line. It differs in that it is a relatively consistent 0.100 thicker from the 12” chord station to the aft most point. The thickness tapers back down from the 12” chord to the leading edge. I have an overlay drawing and others that I will submit whenever we determine what is most useful.

      I’ve done some simple testing of my profiles with paper and mylar and the current paper being used by Duncan-Parnell is surprising stable when cycled in and out of my building but it can’t substitute for the mylar, for the reasons that you pointed out.

      The Dwg 4 mylar is indeed the master that Bob would like for us to use and the one that has been used by all previous LSA builders, AFAIK. But it is, from all my research and measurements, a modified version of the 30-613.5, as briefly described above. Since I started this discussion, and currently, I have no reason to choose the 30-613.5 over the Dwg 4 profile. My decision will be made after learning the intended purpose of the modification or, with your help, the performance characteristics of each profile, combined with your expertise, that would assist in making the decision. We have no test data, demonstrated or otherwise, that would allow a comparison between the two. Should the differences appear negligible, then I would choose the thinner, lighter wing and trust that the load calculations were performed with that airfoil and spar dimensions, as I believe it was.

      The 30-413.5 ordinates were never published but I recently calculated, plotted and linked to the file that, if printed and overlaid with your Dwg 4 mylar, would likely match except for the lower perimeter as it does with the 30-613.5. The profile drawing and ordinates table files for the 30-413.5 are linked in Post# 16. You will also find there, the 30-613.5 drawing profile with all ordinate dimensions included, for your reference if interested.

      I have lost hope that I might soon be enjoying one of your cold beers. Your wagers were also things that I consider to be accurate or, at least, believable. Maybe next time… (insert sad face)

      Best Regards,
      Mitch

    • nborer
      nborer commented
      Editing a comment
      Well, there's one beer back to you... in my reads of the Riblett book, I never noticed that the ordinates on page 71 just above Table I and Figure III-19 didn't follow the format of the previous pages (columns of station, 15% thickness, camber line, and then upper/lower ordinates for 12, 15, and 18% thick airfoils). So it's all right there in terms of the 13.5% airfoil with the 30-6 camber line. I guess I reconstructed it appropriately because my calculations are within 10^-4 accuracy of what he has listed, and I figure we both are just using slightly different criteria for round-off error.

      That said, I suggest we don't go for "net beers" but instead trade rounds. It's more fun that way.

      I've gone through the files you posted... I believe the drawings you have are for the Riblett-defined 30-631.5 airfoil, scaled to a chord length of 60 inches. In your first post are some .dxf and .pdf files and others that I'm unable to download - do those have your as-measured drawing #4 airfoil? That's what I'll need to use for the comparison.

      I also see that you had placed the chord line outside of the airfoil shape. Is that so you can get a blunt trailing edge? I'd recommend scaling the airfoil slightly larger so you end up with a blunt trailing edge in the formblock. Riblett mentions this on page 14, where he says to "cut it off square" at 99% of the chord. Presuming you want to preserve your chord length, just build the template but multiply the x station by (1/0.99 = 1.010101...). That is, scale your drawing as if your chord length were 60.61 (60/0.99) rather than 60 inches. I notice that the blunt trailing edge in my Model B formblock drawing is more equivalent to cutting off at 98% rather than 99%. Note that this may also be a part of the source of the slightly larger thickness, but doesn't account for the whole discrepancy (that is, it'll increase the apparent thickness of the blunt airfoil from 13.5% to 13.63%, or even 13.78% if you go with the 98% rather than 99% chord length). It'll also push back the apparent point of max thickness, camber, etc. by the same relative amount.

      Blunt trailing edges can be hard to implement in vortex panel codes because they need to enforce the Kutta condition - basically, the velocity on the upper and lower surface must be parallel and depart the trailing edge in order for the code to calculate any lift. These codes neglect viscosity and compressibility when calculating lift (and drag), but, paradoxically, no lift and drag is possible without viscosity or compressibility. These terms are very small for low-speed flight, but (viscosity in particular at low speeds) is critically important. If you feel like Googling, look up D'Alembert's paradox. We use inviscid, incompressible assumptions with the Kutta condition in potential-flow codes because they are fast and reasonably accurate (and prior to full-blown CFD, were the best we had), but like all engineering models, they are an approximation of reality.

      In reality, sharp trailing edges are at best unnecessary, and at worst can cause even more problems - both with the flow and especially with structures (and even operations - imagine hitting your head on a razor-sharp trailing edge - it hurts enough with a blunt one!). A little bit of thickness with a reasonably sharp corner acts to force flow separation if it has not happened, sort of "enforcing" the Kutta condition in reality. Also, you can imagine that you quickly get to zero bend radius with trailing edge pieces, which is of course a no-no. So, you can view this as cutting off the theoretical trailing edge from your chord length, as Riblett recommends, or by extending your theoretical airfoil beyond the trailing edge, as I wrote about above. The real question is designer's intent... when Bob put a 65-inch chord line on my Model B drawing 4, that was 65 inches to a blunt trailing edge, and judging by the thickness, it appears to be around 98% of the theoretical chord length - putting the imaginary chord line out to 66.33 inches, if one were to scale airfoil drawings with a sharp trailing edge.

      This thread has caused me to dig back into some work I did almost two years ago, when I had plans but was ordering materials and messing around with formblocks instead of building ribs. Given that my current circumstance makes it impossible to build, it's been fun - I haven't thought about any of this for a while. I even found one of my old large-format scans of drawing 4 (I did this prior to cutting the Mylar to lay over the formblock), and will try to correlate those measurements to my hand measurements. Correcting large-format scans is always a bit of an art... they never come out perfectly perpendicular or scaled, which is why I abandoned using that approach when I first started messing with it. I have a few more tricks that I'll try. It sure beats installing crown moulding... though if I can get that done, and a dozen other jobs, I may finally be able to get my shop back!
      Last edited by nborer; 10-03-2021, 10:18 AM.

  • #19
    How about building an LSA or Patrol with the right wing Riblett and the left wing Bob Bearhawk? Then you can have a really good real world test! : )

    Kevin D
    KCHD

    Comment


    • NCLSA183
      NCLSA183 commented
      Editing a comment
      Now that is an example of why this forum is so important to it's members. So many great and unique ideas are presented. Sadly, I would have absolutely no idea how to connect the test instrument's in this scenario.

      Thanks Kevin!!

  • #20
    Below is a link to the Excel workbook that I created and used while exploring the new and exciting world of airfoils and ordinates. I have enjoyed my foray into this strange world, immensely. I’m sharing this with our members because there may be someone out there in Bearhawk land that enjoys learning new things as much as I. Maybe even the next generation Harry Riblett is lurking in the shadows.

    I spent some time recently reading over the guidelines for “Fair Use” of copyrighted material. Some of our major universities have great information on their websites. After careful thought, I have included two data tables from Mr. Ribletts airfoil book. I had originally planned to only include a small sample which would require that you fill in the rest of the data from YOUR own copy of the book. Although that is no longer required, I would still like to encourage you to purchase a copy if you haven’t already. The Harry Riblett family donated the copyright to the EAA.

    I have included some basic instructions on the ReadMe worksheet that I hope will be enough to get you started. If you have any questions, please don’t hesitate to ask.

    Have fun!


    Mitch


    Riblett_Airfoil_Thickness_Calculator.xlsx
    Last edited by NCLSA183; 10-02-2021, 09:44 PM.

    Comment


    • #21
      Nick,

      That is very considerate of you to cancel the beer wager as we both know that it would be me making all the runs to the 7/11. I’m sure that I discovered the 13.5% ordinates only because it was my single-minded goal. I could not have reconstructed them as you did. My criteria for round-off was just whatever Excel suggested.

      I must take a moment to thank you for your time and patience. You surely have more important things to do than to teach me a foreign language. I will make every effort to ensure that it is not a complete waste of your time. Feel free to just give me Cliffs Notes so I won’t feel as guilty about occupying so much of your time.

      The files from my first post were only left up for a few days because I was questioning myself regarding copyrights, in the beginning. Also, I did some review of them and re-plotted a few different times to gain more experience and try a few different things.

      The chord line was placed outside of the airfoil only because Bob did it that way on Dwg 4. I will attach a scan of the trailing edge of LSA Dwg 4 for reference. I will also spend some time in trying to understand your suggestions regarding the blunt edge and I’m always interested in a paradox of any kind. I once owned a Cessna 172 and I’m just over 6’2” so I have had a love/hate relationship with the pointy end of the ailerons many times. Especially if I happened to be wearing a ball cap that obstructed my vision. Yikes!!

      It pleases me that you are having fun with this. It certainly is fun for me. As I alluded to earlier, don’t let me prevent you from doing the more important things so that you can get back to your shop. I can relate closely to your crown moulding project. I spent my high school summers and then, after the Army, all of my free time during college working as a framing and finish carpenter. My brother had a construction business and I was usually the sawman. Probably because I enjoyed math and I was a bit OCD regarding accuracy. Interior trim work was usually our rainy-day projects. I enjoyed it and the experience allowed me to build my home several years later.

      Thanks,
      Mitch


      This is a scan of the LSA Dwg 4 trailing edge. The tick mark that Bob made near the end of the chord line is the 60” chord station. Maybe you can speculate on the designers intent.
      Dwg4_Aft_Profile.jpg
      This is my most up-to-date DXF of the GA30-613.5 profile. I put the control points on a separate layer so you can hide or show as needed. I hope that feature is universal with other CAD software.
      Riblett_GA30-6135_Base.dxf

      This is my latest DXF of the LSA Dwg 4 with control points. I believe it to be a modified 30-613.5 rather than a completely different airfoil. I experimented with several different thickness distributions in an attempt to find something that matched, but no luck. It just seems to be an added thickness to the true 30-613.5 airfoil profile. You will understand better when seeing the overlay.
      Riblett_GA30-6135_Dwg4.dxf

      This is LSA Dwg 4 overlaid on the true GA30-613.5 profile. The only place it differs is the red lower perimeter line as I mentioned in my last post. This is what it looks like when I place the Dwg 4 mylar over a printout of the 30-613.5 profile. I didn’t include the blunt end, which I measured to be 0.625” back from the 60” chord station. The red line is not completely parallel to the 30-613.5 either by design or just small plotting error.
      Riblett_GA30-6135_Dwg4_Overlay.dxf

      Let me know if you would like these profiles in PDF format.

      Comment


      • #22
        Nick,

        As you are probably aware by now, Bob did as Harry suggested and scaled the airfoil chord to 60” and then squared it off at 99%. He closed off the gap as in the attached drawing, below. Possibly the same as on your BH4B.


        I honestly don’t remember having ever known about the Cooper-Harper ratings but it’s possible that I did at some point. Regardless, I do now. Thanks.

        I had read mention of XFOIL a few weeks ago but haven’t had time to read in-depth or watch a video. Probably more than I want to get involved with, at this time. I’m not planning on trying to design an airfoil but I would like to know more regarding the science and engineering.

        D’Alembert’s paradox will also have to wait a few more days, as well.

        It seems that when I was doing my careful measurements and comparisons of the LSA plans vs the 30-613.5 airfoil profile, a couple month ago, I missed something. It’s still true that I measured and verified that the rib profile and the spars matched with the 30-613.5 airfoil as-drawn. But, a few minutes ago I was flipping though the plans sheets when I realized that I had not measured the components on Dwg 10, Aileron Spar & Rib, although I had measured the rear spar on Dwg 9 but not the aileron spar on Dwg 9. That area was very congested and hard to accurately measure so I skipped it since I had already verified the airfoil itself. My bad. Turns out that Dwg 10 is drawn and dimensioned with reference to the profile of Dwg 4 instead of the 30-613.5 as was used for Dwg 3, 5 and, and part of 9. The rear spar and the pocket rib template do reference the 30-613.5 profile but the 1/2 scale aileron matches the Dwg 4 profile. This can of worms are really wiggling.

        I’m not really sure, at this point, how much that is going to affect/complicate things for me should I decide to build the 30-613.5 wing. Going to need some time to study that area of the plans. Bob’s instruction’s regarding the ribs and spars are basically, build the ribs to Dwg 4 and then bend the spars to match. So, pocket ribs whose template would result in being formed to match the 30-613.5 profile and then fitting to the aileron nose that match the Dwg 4 profile. I suppose the aft radius of the pocket rib fits good enough to the nose of the aileron to have little, if any, effect. Darn worms, stoppit!! Gonna take a good nights sleep to sort this one out…

        Mitch


        img014.jpg

        In this scan, The rear spar and pocket rib are 30-613.5 but the aileron is Dwg 4. Of course, once you build rib and spar to Dwg 4, the rear spar will be 0.100 thicker, or more.
        img015.jpg
        Last edited by NCLSA183; 10-05-2021, 11:22 PM. Reason: Added scan of aileron and additional text

        Comment


        • Frank
          Frank commented
          Editing a comment
          I'm not up to speed with you guys by any means but I have been pondering what might be the ramifications if you depart from drawing 4, the Mylar. If it were me I would be worrying about unforeseen consequences. Also (and I may be missing the point regarding your aileron issue) my understanding is the aileron has its own airfoil shape independent of the wing and the two surfaces are not "flush". Also, when forming my pocket ribs I left the single flange "tab" unbent so that I could bend it to fit the rear spar. Just a few comments from the peanut gallery. Carry on, gentlemen!

        • NCLSA183
          NCLSA183 commented
          Editing a comment
          Hey Frank. I’ll never be up-to-speed with Nick but I’m having fun accelerating. My brain lost its ability to impersonate a sponge many years ago. But, I am doing my best to avoid anything unforeseen and I welcome your input and advice on things that I might need to avoid or to give special consideration.

          It is also my understanding that the aileron is a different airfoil and I had verified that it was different than the aft section of the wing profile. I incorrectly implied in my previous post that the aileron matched the Dwg 4 profile when I only meant that it matched in regard to the Dwg 4 aileron spar height and then, based upon that, the aileron was designed to be a thicker airfoil to complement the Dwg 4 wing profile rather than the thinner 30-613.5 profile. I’m only trying to relate what I see and measure into, hopefully, useful information for Nick. Thank you for noticing that my explanation was clear as mud, maybe still so. LOL. When it comes to understanding airfoils, I AM the peanut gallery.

          When I was posting Tuesday evening, I had wondered to myself, what were builders doing to make the pocket ribs fit the thicker Dwg 4 rear spar. I like your simple solution. At what point were you aware that the full scale, pocket rib template would only fit the 30-613.5 spar and not accurately fit the Dwg 4 rear spar that you were building/had built? Was it something you had read on the forum or after you had formed the inaccurate pocket ribs? These discrepancies on Dwg 9 combined with the others that I’ve brought up in this discussion has only reinforced my stated belief that the modification of the 30-613.5 profile was an afterthought in Bob’s design process. At what point and for what reason, I can’t say without speculation. Only the reason is of interest to me.

          Mitch

        • Frank
          Frank commented
          Editing a comment
          Hi Mitch, my thinking on the pocket ribs was this: the center and back ribs are formed and the spars are bent to fit; but it might be best to fit the pocket ribs to the rear spar after, not before. Not noting any "discrepancy" in the plans but noting the general "instruction" to make the parts fit to each other. Maybe I will succeed.

      • #23
        Mitch:

        Thanks for the updates. I haven't had a chance to import the new drawings and finalize my comparisons; that could take a bit as it's a busy few weeks here (I'm treasurer for my kids' PTA, and it's tax form season, so that means my spare evening cycles are taken up with PTA stuff). I did resurrect my smoothing codes that I wrote in Python; long story, but the only reason I code in Python for personal projects is that the stuff I use for work requires licenses that cost way too much, and I'm not allowed to double-dip on those machines. So, I'm re-learning how I implemented the airfoil smoothing technique I used when I did these computations 22 months ago from raw data. I use Python a little, but I don't use it for work, so I'm a little rusty at the moment. I'll get there.

        Why smooth, you ask? Any numerical analysis, in particular the vortex panel code I'm using (and the predecessor code that Riblett used) really don't like noise (based on our measurement tolerances) in the inputs - this noise appears as artificial variations when solving the system of equations used to get the answers for lift, drag, etc. In "real life," those little variations (like flush rivet dimples) are, essentially, "filled" with stagnated air, but in an inviscid panel code, there is nothing to "fill" it, so it introduces numerical noise that looks like variations and lift and drag that aren't physical. Riblett actually alludes to this in his book; in the first paragraph on page 12 (6th ed.) he laments some issues with the "improved" version of the Eppler code, including "the results appear to be overly sensitive to data point (ordinate) accuracy, thus I still favor the old program." When plotting Riblett's raw ordinates, I can see that the data in the book does not appear to be smooth to the precision that most numerical solvers love to see. That doesn't make the airfoils bad; it just makes the numerical predictions of their performance challenging, especially as you use more accurate methods. My current process smooths the airfoils by fitting the points to functions from Kulfan's CST method; I won't belabor them here but if you want a heady read here ya go: https://brendakulfan.com/docs/CST2.pdf. The short version is that it makes airfoil ordinates smooth in ways that make numerical solvers behave better.

        My plan is to try to extract the points from your Drawing 4 .dxf file by first converting it over to an image and then sending it over to one of my data grabber programs - I use Engauge or more recently, Graph Grabber. That process itself introduces error. Since you have the native CAD file, I could eliminate this step if you could send an Excel or even just a CSV or some other delimited text file of the drawing x, y coordinates for the airfoil ordinates and chord line (I don't have AutoCAD, so I'll need to download a translator to get the .dxf as an image). If not, no worries, it doesn't take much longer... it's just a step that introduces a little more error.

        Once I have them, and I have smoothed ordinates, it'll be pretty straightforward to run the cases. I've already run the Riblett-defined 30-613.5 airfoil in XFOIL, so it's a matter of getting the other airfoil from drawing 4 into the system. I've also run a few validation cases, since a few folks have (rightly) called into question the accuracy of what I've been presenting. I found some good data on the NACA 4412 and 4415 airfoils from wind tunnel tests (Abbot & von Doenhoff and NACA report 563, though the latter does not include drag information) that I digitized as "truth" data and ran against XFOIL. I'll present those validation cases as well so we can get an understanding of the tool accuracy. I'm including a teaser of the validation data below for the NACA 4412 at a Reynolds number of 3 million - not awesome, but certainly pretty good, at least until we get near stall.

        XFOIL_comparsion_NACA_4412.png
        4-Place Model 'B' Serial 1529B (with many years to go...)

        Comment


        • #24
          Nick,

          If I wasn’t completely clear in a previous post, then please hear this… Post #21 and #22 were only meant to provide you with information that may assist you in some way with this “project”. I do not require or expect a prompt response. If you would like to continue working with this, then please only do so when and if you have the time and desire. I still plan to talk to Bob at some point before starting my project with the hope of learning his design goal for the airfoil modification. The information and experience that you are providing will go a long way in helping with my decision and the things that I am learning on this journey will always be invaluable regardless of what decision I make. In order to get started building in this calendar year, my backup plan has always been Dwg 4.

          I think Python is great and I’ve dabbled a bit over the years. I read recently that it is rated #1 this year by the IEEE. Most of my client work the past 5-10 years have been either VBA or C++ (as used with the Arduino line of hardware and clones) and a hodgepodge of others languages over the years, including low-level assembly on an 8748H microcontroller circa 1991, I think. Yeah, you’re right… that was TMI. At least I didn't go back into more ancient history, so consider yourself lucky!

          Yes, smoothing. You read my mind. LOL. More acceleration!! (referring to my reply to Frank) I can actually visualize those lazy, stagnated molecule’s being shooed away by those geeky numerical noisy molecule’s that refuse to get physical.

          I wasn’t sure what all you needed from me so I apologize for leaving you short of data. I’m sure that is considered to be cruel and unusual punishment to an engineer. So, to brighten your day, I am attaching an .xlsx with the ordinates for Dwg 4 in the Riblett format. If I understand correctly, you won’t need an image file as long as you have the ordinates. But if needed, I can supply most image formats or a PDF. I haven’t used AutoCAD since the late ‘90’s but have gotten by with free or opensource alternatives until recently when I tested and purchased QCAD Pro for $39.00, (available free with limitations). I think it will serve me well in the foreseeable future.

          As always, I’m looking forward to reading, or at least trying to read, the Kulfan’s CST method and learning the results and data from the cases mentioned in your last paragraph.

          Good luck with the tax forms. Always been my most favorite part of life…

          Mitch



          Ordinates data in Riblett style format
          Bearhawk_LSA_Dwg4_Ordinates.xlsx

          DAT file for importing control points into QCAD
          Bearhawk_LSA_Dwg4_DAT.txt

          Comment


          • #25
            Mitch:

            Thanks for the ordinates. I just took them without 3rd party smoothing and just used XFOIL's smoothing routine (the PANE command, if you're deep into XFOIL menus and you want to make it work). So, these results are without Kulfan smoothing, and a comparison of the Riblett 30-613.5 airfoil to the LSA drawing 4 coordinates you provided.

            Looking at the ordinates, I see that the lower surface is translated slightly downward in the LSA drawing 4 vs. the Riblett coordinates. I believe this is to help with the blunt trailing edge. I still see some waviness in the lower surface that doesn't seem intuitive, but that's present in both forms - the LSA drawing and Riblett coordinates.

            Overall... both airfoils perform very similarly. Not surprisingly, the slightly thicker drawing #4 airfoil with the thick trailing edge does slightly better at higher angles of attack. That may be an artifact of the analysis code more than anything else. From the validation results, I would call this performance "essentially identical" if I were giving a design review.

            I'm biased, but I say go with drawing 4. You'll be in good company with other LSA builders, and the performance difference is at worst nil, and at best my eek out a small amount more of high-lift (though I don't really believe those results, for reasons I won't belabor).

            Build away, my friend!

            Low-speed comparison

            LSA_comparison_Re_3e6.png
            High-speed comparison
            LSA_comparison_Re_6e6.png
            Geometry comparison - scale and labels distorted for copyright reasons
            LSA_vs_30-613p5_geom.png
            Nick
            4-Place Model 'B' Serial 1529B (with many years to go...)

            Comment


            • NCLSA183
              NCLSA183 commented
              Editing a comment
              Nick, so sorry for being AWOL for a few weeks! My dang ducks would not stay in a neat row as I had instructed them. A couple of them just had to get out-of-line and take up most of my free time the past few weeks. I am, of course, referring to clients. My plan was to take down my shingle at years end but that date might slip a bit. As may my LSA start date. But I still have hope.

              I suspect, that after you read my post #22, you likely saw the writing on my wall. In the beginning, I either stated, or alluded to, the fact that I had no intentions of attempting to redesign, reengineer or otherwise modify the design of the LSA. I couldn’t if I wanted. I only wanted it to be a true Riblett 30-613.5, as the plans partially were and as it was advertised. So, when I discovered that I had not verified the aileron dimensions on Dwg 10, and subsequently learned that it matched with Dwg 4 instead of Dwg’s 3, 5 & 9, then my quest to build a true Riblett 30-613.5 wing was over. Even if I could, I had/have no desire to try designing and building a new aileron that would complement the 30-613.5 wing. Case closed…

              What have I learned from all this? Well, quite frankly, some things that I am greatly disappointed with. Was this only an exercise in futility? Depends on one’s outlook. Was it avoidable? Yes. But I don’t regret the time and knowledge that I gained. And, I truly enjoyed the interaction with the members of this forum. What a great bunch of generous and knowledgeable people. Thanks to you all.

              Nick, again, thank you for your patience, expertise and, most of all, your time. I can’t say that I’m surprised by the results of your comparisons but it’s really nice to see it confirmed by your analysis and calculation’s. And, of course, I always knew that you were biased, from the git-go. LOL Who wouldn’t be! You also understood where I was coming from with my questions and curiosity and, instead of writing me off, you made extreme efforts to help me understand and to answer my questions. You are, mi amigo, a great asset to this forum and will always have my sincere appreciation and respect.

              Sincere regards,
              Mitch

          • #26
            Fellas, I am not expert or particularly knowledgeable about aerodynamics but I like to think I am open minded enough to never look a gift horse in the mouth. I chose the LSA for many reasons, among them is that it is an easy airplane to fly, it uses a relatively inexpensive engine, it has great performance. It ticked all the boxes that were important for me. The type of airfoil type is a moot point for me. So I wanted to bring up a point about the designer of the air foil that not many people may consider. What was the motivation of the designer in the first place. I am in no way saying or implying here that this airfoil is unsafe because it has proven its self to be perfectly safe and a high performance airfoil. I only wanted to share a story about several designers, I believe it is a very insightful look at the motivations that make some people claim to be experts in their chosen fields and I wanted to share it with you guys just as food for thought. If you read William Wynne's website you have probably already read this article. If not and you would like a little more insight to Mr. Ribletts motivations please read this article. I offer it for your edification. https://flycorvair.net/2013/01/28/ex...iation-trials/

            Comment


            • nborer
              nborer commented
              Editing a comment
              Insightful piece. I’ve been an EAA member since 2003 but didn’t actually read about Riblett much prior to getting my Bearhawk plans. After I got my plans, I bought his book and was… shall we say surprised with the tone. I agree with many of the principles, but don’t buy the conspiracies. I currently work with the designer of a few of the airfoils that Riblett disparages, and he (and his wife) are among the sharpest aerodynamicists I know. They didn’t “screw up” - their airfoils do exactly what they’re supposed to do.

              My basic message in this thread, which I’ve tried to back up with analysis, is that there is likely little if any perceptible difference in airfoil performance between whatever profile Riblett published and the one that exists in the Mylar drawings.

            • Frank
              Frank commented
              Editing a comment
              Interesting reading, Dave. My takeaway is there is a danger in knowing too much (but not everything) about a topic, or about people in general. Once the door is opened a crack on various personalities it's hard to close it again. I agree with you: it's the LSA specs (and esthetic appeal) which brings me hereI​​​​​​​​​​​​​​

            • NCLSA183
              NCLSA183 commented
              Editing a comment
              Thank you Dave. I enjoyed the read.

              Mitch

          • #27
            HI Frank and nborer please don't get me wrong, I am a fan of the airfoil on the LSA regardless of who designed it but I like to play devil advocate in most discussions i participate in and try to bring up several points of view. My friend William seems to be more than just an engine builder but a philosopher and he puts ideas into words that I would have a difficult time doing. I totally agree with his philosophy and his out look on life and the way we treat other people. I think it is a fair philosophy and most of all very objective based on experience and empirical data. . I believe William is a guy who would never sell out his morals and principles for profit. Im my personal dealing with him he has never done me wrong and has been more than generous to me and always treated me with the utmost respect.

            Comment


            • nborer
              nborer commented
              Editing a comment
              davsLSA: I'm in agreement with you and with the article you posted, at least about the Riblett discussion. When I said "I don't buy the conspiracies" in my post above, I was referring to the ones in Riblett's book about NACA, criminal negligence, etc. - not Mr. Wynne's article. Sorry if I didn't make that clearer. I'm actually in full agreement with Mr. Wynne - the aerodynamic principals Riblett proposes in his book for "good" airfoils are, in fact, good, but otherwise the tone of "GA Airfoils" is a bit dramatic and overstated. I've tried to indicate that in my posts, but also don't wish to offend, since I do know that many people think very highly of Mr. Riblett. Dave Lednicer provides a much less-nuanced takedown here: https://yarchive.net/air/airfoils_riblett.html

              The introduction to "GA Airfoils" starts by stating incredulity that NACA airfoils were not designed for actual airplane use, but rather were wind tunnel test specimens used to investigate effects of different geometry changes on airfoil performance. This was the first thing I learned in my aerodynamics class when we were introduced to NACA airfoils and the seminal tome of Abbott & von Doenhoff. Yes, people have used those airfoils in a variety of designs, but the intent of the research wasn't to create a catalog of airfoils - it was to investigate different geometries. Maybe it's not well-known outside of aerospace engineering, but it's common knowledge in the field.

              Another "surprise" for some folks is that airfoils actually don't matter a whole lot on their own - they need to be mated with an appropriate 3D wing design and tail planform. A "good" airfoil can perform very poorly on a "bad" wing and tail, and vice versa. I learned this firsthand, at least computationally, when I started down the path of rotor and propeller design several years ago. I randomly generated thousands of airfoil shapes (using the Kulfan method linked above, so they would be "airfoil-shaped") and created minimum-induced loss propellers (an analytical method for optimum chord and twist distribution for a "design point" - a specified thrust/power/torque at a specified RPM and airspeed) with those airfoils. There were a lot of dogs, but what surprised me were how many of them got with 1-2% of the maximum propeller efficiency seen in the whole lot. The airfoil design mattered far more for good off-design performance than for best on-design performance. So, pick your favorite airfoil, then tailor it to a wing and tail planform to it for good performance. That's what Bob has done here with the LSA and others.

              Overall, I'm trying to help the OP make a choice about what shape he uses... with gentle nudges towards the as-drawn #4 profile on Mylar. But it's up to him. From an airfoil performance perspective, my admittedly simplified analysis tools indicate very little, if any, performance difference.

              Nick

          • #28
            HI Nick, Your posts are very educational and I enjoy reading them, As any pilot the study of aerodynamic is a subject I'm always interested in and knowing more about. You are certainly not offending me with your posts. And besides God knows I could use more education.
            Last edited by davzLSA; 10-11-2021, 10:27 PM.

            Comment


            • #29
              Would it be illuminating to this question to derive the coordinates of the ribs being produced and supplied with the LSA Kit? Perhaps someone here with the kit could take sufficient measurements of a rib (and the associated other components that connect to the rib if need be) to deduce if the kit follows the mylar rib shape (drawing 4 rib shape) or if it follows Riblett's published coordinates.

              It seems that this information could help LSA scratch builders who encounter this apparent discrepancy in the plans and find it problematic for their project.

              Comment


              • Mark Goldberg
                Mark Goldberg commented
                Editing a comment
                The ribs in the kit conform to the mylar. MG

              • NCLSA183
                NCLSA183 commented
                Editing a comment
                Thank you, v.x, for your comment. You succeeded in getting an answer to a question some of us have asked but failed to obtain. It does confirm my thoughts from an earlier post.

                Discrepancy, like beauty, appears to be in the eyes of the beholder. It doesn’t seem that the designer or the kit manufacturer consider it to be a discrepancy, or at least one worth noting. Possibly because it ##doesn’t affect the safety## of this aircraft. By definition, the differences that I have outlined in this conversation are, indeed, discrepancies. Subjectively speaking, the last place that I want to find discrepancies would be in the plans for an airplane that I was building and/or flying, regardless of its proven flight record. And, any unexplained and undocumented discrepancy, as is in the LSA plans, would, in my opinion, be problematic enough to warrant the questions that I have presented in this topic. NOT problematic in regard to any safety reasons. I have never said or meant to imply such. But rather, problematic, as defined in being “difficult to solve or decide”. My reasons are outlined throughout this conversation.

            • #30
              Originally posted by v.x View Post
              Would it be illuminating to this question to derive the coordinates of the ribs being produced and supplied with the LSA Kit? Perhaps someone here with the kit could take sufficient measurements of a rib (and the associated other components that connect to the rib if need be) to deduce if the kit follows the mylar rib shape (drawing 4 rib shape) or if it follows Riblett's published coordinates.

              It seems that this information could help LSA scratch builders who encounter this apparent discrepancy in the plans and find it problematic for their project.
              There is no need for this to be problematic problem. I have enjoyed this discussion from the academic perspective but realize the drawback of hosting it may be confusion about whether there is something wrong with the airplane. I hope that folks won't arrive at that conclusion. Build the airplane to match the plans and you will have a strong and light, fantastic, proven performer.

              Comment


              • NCLSA183
                NCLSA183 commented
                Editing a comment
                Regardless of how carefully words may be chosen, there is always room for personal interpretation. A reader should not hesitate to question any statement that may seem vague, misleading or differ significantly from generally accepted norms. I hope, as well, that arriving at an unintentional conclusion was not the case in this conversation.

                I respectfully suggest that a bit less vagueness from the designer would be greatly appreciated by those who wish to understand the reasons and rational behind mid-design modifications and especially in regards to drawings that don't factually represent the designers published specifications. This has been somewhat problematic for years without resolution. Unless, of course, one was sufficiently openminded and receptive to data from sources other than the designer.(wink)

                Thanks for allowing this topic. Ironically, and in reference to my original post, it is I that wishes for a refund of his time...
                Mitch
            Working...
            X