Learn from my mistake Design flaw

[Flygirl1]

This is a photo of a design flaw. Do not use 9 #3 squeeze rivits to hold you cowl doors on, they can fail. You need at least 20 #3's here!! 5 of the 9 sheared off in flight in quick order. Why the last 4 held, I don't know. Doing about 139 MPH at the time and just prior to that doing about 144 indicated in a decent. These rivits still had the head held in place by paint and the stem was still in the hole. I guess other than the fact it happened at all, it was my lucky day they didn't all fail! Not sure what would have happebpned to my cowl door and cowling, but I'm sure I wouldn't have liked it! Our plan is to replace the first in line with a screw and nut plate and add more rivits.

[Chris In Milwaukee]

Wow, glad everything held together!

[jaredyates]

In the 4-Place we use a steel doubler at the front and back of the hinge. Does the Patrol use anything similar?

[Flygirl1]

The plans book calls for 9 #8 screws and nuts. In that area.

[Battson]

This is a photo of a design flaw. Do not use #3 squeeze rivits to hold you cowl doors on, they can fail. 5 of the 9 sheared off in flight in quick order. Why the last 4 held, I don't know. Doing about 139 MPH at the time and just prior to that doing about 144 indicated in a decent. These rivits still had the head held in place by paint and the stem was still in the hole. I guess other than the fact it happened at all, it was my lucky day they didn't all fail! Not sure what would have happebpned to my cowl door and cowling, but I'm sure I wouldn't have liked it! The flaw came in either I needed more #3's or larger rivits. Our plan is to replace the first in line with a screw and nut plate and replace the rest with larger rivits.

Thanks for sharing this - a great opportunity to learn. I bet that was very frightening when it happened, if it totally unzipped it would have been a very bad day.

I don't think there's anything wrong with the rivet diameter you chose. That size is set by the metal thickness you're riveting together. Using a larger river just means the substrate fails before the rivet does - but it will probably still fail.

It's the number of rivets which I would focus on. If I understand your post correctly, there were only nine (9) rivets holding the door on?!
I think AC43.13 suggests a panel that size needs at least twice that many. I think the recommendation is rivets at 1.5 inch centres, from memory...?

Several rivets have cracked in that area over the last 550hrs flight for us. It's a high vibration area, perfect conditions for fatigue. There are somewhere between 20 or 30 rivets holding it together, so we weren't in any hurry to replace cracked rivets.

[Flygirl1]

Actually it wasn't too bad, I was too busy-- spent the next (slow)10 miles to the airport praying it would hold. ;-) I hated to admit it but it was my mistake and after reading your post, thank you for posting, I agree, more rivets are needed here. If I can keep that kind of stuff from happening to somebody else I'll be first in line to admit my mistakes!! D.

[Flygirl1]

Jared, I do not have the plans in front of me and I'm not sure what it calls for. This is our design flaw and not a flaw on the plans. Having said that, it is a doubler. Studying the photo it looks to me like those holes are too far apart. Had there of been more of them I doubt they would have sheared off like dominos. I feel confident our fix will hold. I'm going to put a #6 screw in the forward hole, or maybe even an 8 or a 10 and some JB Weld, whatever I need to feel good about this!! And I'm doing it to the other side too. Took the fun out of the flight home I can tell you that!! ( put screws and nuts in the holes for the flight home)

[Bdflies]

Yeoowww!! Thanks for posting that. I've got some checking to do.

Bill

[Flygirl1]

Hi Battson, as you can imagine I have been thinking about what happened a LOT. After getting into the repair we noticed that the rivets look more like the heads were “popped” off, not sheared off. Also there were 6 not 5 that failed. And even though it doesn’t make it OK, there is a bracket from the cowling to the boot cowl that is screwed in place that also holds it all together. As far as seeing any cracks there were none, not even in the first one, or in the paint. It’s like the internal pressure was just too much for those rivets. I'm thinking the internal pressure had more to do with it also because if the external air flow was a big factor I think that cowl door would have continued to zip. Once it opened, the internal pressure wasn't as great? No doubt more would have maybe kept it from zipping, but I’m concerned about the way these heads just popped off. I have 170hrs on the plane and it most certainly could have been cracked, but I would never have seen it. I get why the ones following the first one went, but it’s the first one that bothers me. You said you have seen cracks in some of your rivets and I was wondering where are the cracks located in the rivet? I have grounded my plane and am doing a thorough inspection of ALL the rivets and other than paint cracking and or black rub marks from loose rivets, I’m not real sure where and what to look at. Also I’m wondering what kind of rivets you used in that area, round headed or counter sunk. It seems to me that counter sunk would be sturdier due to the fact you have a dimpled metal helping to add to the strength of the head of the rivet?? The only cracks I have ever seen was if a counter sunk rivet got over squeezed and a small crack in the thin edge of the head would appear.

[Bdflies]

170 hrs!?!

Bill

[Flygirl1]

Yea, I know!!  I just LOVE flying this plane!!!

[fairchild]

I kind of figured that the idea of the rivet formula in 44-13 was taking the load and dividing by the # of rivets and each rivets strength-
But i also think of it too- as all of those rivet heads head area's added together give so many square inches. If you use fewer big rivets- the skin tears but the rivets dont fail.
So using many smaller rivets gives enough total head area so the load is spread out more and thre skin doesnt rupture arround the head.
Maybe layer the sheets in steps and use LOTS of small rivets. (like cessna does on the fuselage skin as it comes forward towards the firewall)

be a while before I get to that spot--- but when I do i will take care to build strong !
Baffles and cowls and hoods seem to have the most severe vibration environment of the whole plane.
T

[Flygirl1]

Actually it wasn’t the door skin where the rivets failed but the angle piece the door skin is attached to. I do not know the formula, but I’m sure there is one that also adds the thickness of the material being held by the rivets into consideration. We've changed more than one thing in the way we've attached the door. It WON"T happen again!!

[Battson]

Hi Battson, as you can imagine I have been thinking about what happened a LOT. After getting into the repair we noticed that the rivets look more like the heads were “popped” off, not sheared off. Also there were 6 not 5 that failed. And even though it doesn’t make it OK, there is a bracket from the cowling to the boot cowl that is screwed in place that also holds it all together. As far as seeing any cracks there were none, not even in the first one, or in the paint. It’s like the internal pressure was just too much for those rivets. I'm thinking the internal pressure had more to do with it also because if the external air flow was a big factor I think that cowl door would have continued to zip. Once it opened, the internal pressure wasn't as great? No doubt more would have maybe kept it from zipping, but I’m concerned about the way these heads just popped off. I have 170hrs on the plane and it most certainly could have been cracked, but I would never have seen it. I get why the ones following the first one went, but it’s the first one that bothers me. You said you have seen cracks in some of your rivets and I was wondering where are the cracks located in the rivet? I have grounded my plane and am doing a thorough inspection of ALL the rivets and other than paint cracking and or black rub marks from loose rivets, I’m not real sure where and what to look at. Also I’m wondering what kind of rivets you used in that area, round headed or counter sunk. It seems to me that counter sunk would be sturdier due to the fact you have a dimpled metal helping to add to the strength of the head of the rivet?? The only cracks I have ever seen was if a counter sunk rivet got over squeezed and a small crack in the thin edge of the head would appear.

What do you mean by "popped" off? Did they pull out with the rivet intact?

Normally they crack just under the head, across the shaft where the head tapers in to meet the shank, and the rivet's head falls off.

If they were torn away by great pressure, then you would see a dull lumpy surface on the rivet's torn face. If they cracked by cyclic fatigue, the remaining part of the rivet would have a surface which is part smooth and shiny, and part dull and lumpy.

Both kinds of rivet should be strong enough, provided you've used enough of them.

If you were economical with your rivets and screws during the build, be sure the look at the wings. Both rivets and screws are structural components and you cannot skimp there. The minimum spacing should be about 1.5" for rivets, about 2.5" or 3" for screws. This includes the little backing plates which hold the inspection panels on. Those are structural parts of the wing.

[Nev]

I’ve inadvertently used (13) -4 rivets. Not based on anything scientific or rule based. I wasn’t able to find it in the plans. That’s just under 3” spacing, and the first 3 are 1.5” spacing. Additionally I’ve got a machine screw at the firewall end instead of a rivet. Is this sufficient or have I underdone it ?

[svyolo]

If you look at the first picture, it is "unzipping". Failing one at a time, and probably failing faster and faster as more cowling is exposed to the airflow.

Sometimes, depending on the failure mode, only the next fastener matters. I think seaplane doors are like that. Make the outboard hinge strong enough, and the ones in the middle aren't really doing any work.