Load Test Stub Spar

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

proppastie

Well-Known Member
Joined
Feb 19, 2012
Messages
5,674
Location
NJ
I do not know of a way to test all of the elements of root piece of a spar with a simple load test because the moment on the top/bottom “cap tubes” is derived from a 22 ft. long semi span wing with approximately 43 uprights and cross-members where as there are only 3 uprights and and 2 cross-members on the stub-spar test piece. I designed the test to stress the uprights and cross-members because these are my original design and not a copy of anything I ever saw..... I used a simple truss program to model the whole spar and picked a weight that would test the axle loads and moments of the uprights and cross-members only....I felt comfortable doing this because the loads that tubes were capable of carrying were easy to look up on charts....(ref. Bruhn 75)

I had calculated that 900 lb at a certain location would do the trick and designed the fixture for that.

The plan did not survive the first test because of my in-experience in aircraft design and stress.

First picture shows the original design setup

Second picture:Looking for 900 lb I saw the flat portions of the upright over the jack flexing and I stopped the test at 450 lb....the ratio of the pressure to lb is 2.25 with the 200 reading being 450 lb....
The flat portions were not modeled in the truss program and the truss program assumed elements connected at the centroid of the elements which was not the case....which is why I knew a physical test was needed.

I turned the fixture over and pulled the bottom tube and the cross-members were buckling at again 200 or 450 lb....this would have been a negative G loading and as the spec for negative Gs is ½ that of the positive G loading I was not as upset as I was with the first test.

Without stopping to re-calculate and very upset I decided to pull the positive load but I could not pull from the calculated location so I decided to just do it and see where it ends up.....Every cross-member was in tension and I figured it was worth a try......when I got to 400 reading or 900 lb I decided to pull until 500 or 1125 lb because the location of the pull was inboard of the calculated location...at approximately 460 or 1035 lb the cross-members failed in tension.

The cross-members are easily replaced.....calculated cross section area of item in tension at the failure point fell right on the numbers as regards Ftu or 59000 lb/sq-in for 2024 t-3....which shows that my truss program was correct at least with that element. I really missed it as regards the size of the attachment surfaces of the cross-members but that is easy to fix. I plan on redesign and replacing the cross-members.

My current calculated loads based on the test shows that my cross-members and uprights will survive a limit positive load.... although the current design cross-members would not survive much more than limit negative load.

IMG_20220704_092042.jpg IMG_20220704_092712.jpg IMG_20220705_102557.jpg IMG_20220705_103921.jpg IMG_20220705_104008.jpg IMG_20220705_104343.jpg
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
17,348
Location
Memphis, TN
Right off the bat, the piece that snapped has manufacturing issues. I can’t tell if it’s the grain or shrinker stretcher marks, but those are stress risers and it’s going to break there. That is just as important as the engineering, especially as close as you are cutting the line.
 

proppastie

Well-Known Member
Joined
Feb 19, 2012
Messages
5,674
Location
NJ
calculated cross section area of item in tension at the failure point fell right on the numbers as regards Ftu or 59000 lb/sq-in for 2024 t-3....
1657212234615.png
the cross section through the rivet holes is the width of the item minus the dia. of the rivet holes.....it failed there because that is where the least amount of metal in tension was......that is not to say the rivets might have been overdriven or I was not lined up with the grain......however .7 percent (less than 1%) theoretical difference is pretty close.

here is a better picture of the first cross member that failed ....right through the 2 rivets.
 

Attachments

  • IMG_20220705_104346.jpg
    IMG_20220705_104346.jpg
    46.4 KB · Views: 1
Last edited:

WonderousMountain

Well-Known Member
Joined
Apr 10, 2010
Messages
2,750
Location
Bellingham, Wa
A good test;

Tff is right, orient metal grain lengthwise so it's stress riser is not normal to strain. That said, your diagonals are going to carry near 50% more force under load, they appear to be the same manufacture as the verticals. On the bright side your tube cuffs are holding up well.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
17,348
Location
Memphis, TN
If the numbers are working, good.

Technically the rivets are supposed to pull out before the hard parts break, design wise. The drag brace is not supposed to break like that. The chance it should break in the center should be equal as the chance at the ends. It might have broken with your numbers on a fresh part, but it was not going to withstand multiple cycles of a lesser load.

My point is more along the lines that it was going to break there. Looking at airplane parts without numbers, just about anyone who builds stuff is going to be saying in their head, that’s going at any time.
 
Top