Load Test Question

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
I am not familiar with Cauchy stress. What other options for stress plots do you have ? Are you able to plot the inplane shear stress and the inplane direct stress in the spanwise direction ? That would be more useful .....

I would be careful relying on FEA results in this area of the structure, expecially as we know there is buckling occuring, unless you are running the nonlinear solver and you have included geometric nonlinearities in the solution. This is the only way that the buckling behaviour of that D box skin and flanges can be calculated .... and you will need a very fine mesh to get an accurate result, particularly as the leading edge skin will buckle on the compression side at low stress levels and you will need to continue the solution into the post buckling regime. If you are running the linear static solver remember that the analysis is based only on the initial geometry and cannot predict buckling and cannot calculate the load redistribution that occurs at loads higher than that at which the buckling occurs. A linear buckling solver can predict the load at which the buckling first occurs but is no use here because we are interested in the post buckling strength.
 
Last edited:

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
There might be more web shear deflection than expected because of the tractor mount up? In flight the wing shear transfers directly into the opposite wing.
Mark Calder had a D-cell test failure in a similar set up on his blog.
looking at pictures you should see a large piece of angle attached from the rear fuselage tube to tractor to simulate the attachment of the other drag spar to the other wing.
 

Attachments

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
I would be careful relying on FEA results
yes...... that is why I am attempting calculations my spread sheets demonstrate.....The results are consistent......
I would add it is so very helpful to have an experienced aircraft stress expert checking the calculations and assumptions. Thank you very much.
 

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
There might be more web shear deflection than expected because of the tractor mount up? In flight the wing shear transfers directly into the opposite wing.
Mark Calder had a D-cell test failure in a similar set up on his blog.
Yes .... good point. Do you have a link to Mark Calders page that documents his failure ?
 

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
yes...... that is why I am attempting calculations my spread sheets demonstrate.....The results are consistent......
That is good practise ... too many people have blind faith in FEA without really understanding what they are doing. I have seen some monumental stuffups over the years with FEA.

In my post above on FEA I edited it to add an extra paragraph at the beginning. Not sure if you saw that but I was asking what other plots you can produce ..... some of them might help our understanding of what is going, despite the limitations of the solution. What FEA software are you using ?
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
Are you able to plot the inplane shear stress and the inplane direct stress in the spanwise direction ? That would be more useful .....
The program is very powerful ....cloud based out of Germany. 3000 hr free...I am still using Autocad 14 so I have to load into Fusion 360 then save a Catia and upload very...... cumbersome for me but I have no desire to try to learn SW or any other newer software, as I a running out of time. 72 1/2 ....

Cauchy stress magnitude
Cauchy stress Point
x displacement
y displacement
z displacement
all displacement
total strain magnitude
total strain EPXX, EPYY,EPZZ,EPXY,EPXZ,EPYZ
Von Mises
Family node
family cell
numldcell

cutting planes
isosurfaces
isovolumes
particle traces

I wish I knew how to use or what all these are for.......very difficult in a vacuum.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
if you want to play with it you can download/modify my projects as they have to be "public" for 3000hr free
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
What FEA software are you using ?
currently used 110 core hours out of 3000
 
Last edited:

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
Are you able to plot the inplane shear stress and the inplane direct stress in the spanwise direction
not sure what exactly you mean.....however the spar was constrained in the airplane x (fore and aft) so the numbers you are seeing are only in the airplane Z direction (up and down) Looks to me that Cauchy and Von Mises are about the same. ( give very similar results).....
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
Where I am at today with your help,.....given the Leading Edge has no contribution past its buckling point the Spar is carrying all the loads.

Looking at the 40% load case for the Spar. MY/I = 6900 psi. at the top Spar/Leading edge assembly.

Looking at the Limit load case for same the MY/I= 17,358 psi

1593823958836.png

Looking at Crippling for a flat sheet .016 x 18 (LE length in the X direction) (see Spread Sheet) or .016 x 21 (span between ribs) I get 7400 PSI and 7000 Psi.

Looking at Crippling for a flat sheet .016 x 2, .036 x 1 I calculate crippling at around 17K psi (see Spread Sheet)

Are these flat sheet calculation valid.? The fact we see buckling at the 40% load says that calculation for the flat sheet is valid,

This is where experience really counts.
 

Attachments

Last edited:

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
not sure what exactly you mean.....however the spar was constrained in the airplane x (fore and aft) so the numbers you are seeing are only in the airplane Z direction (up and down) Looks to me that Cauchy and Von Mises are about the same. ( give very similar results).....
Most FEA software that I have used allow the components of stress to be plotted along the local axis directions of the plate elements (x,y) and the global X, Y, Z axes of the model. Also the shear stress (e.g. xy in plate axis or XY/YZ/XZ in the global system) can generally be plotted. In your D box model, it would be beneficial to be able to plot the component of stress along the spanwise direction and the shear stress to be able to study how the bending stress in the D box skin is sheared back into the spar as the root rib is approached. However, it does not look like you have the ability to do this in the software you are using. The Von Mises stress is a combined stress which is basically an equivalent stress (i.e. one value that represents the more complex state of 3 components of direct stress and 3 components of shear). That is great if you wish to see whether something has yielded but it does not give any information on the direction of the stress or what components of stress are contributing so it is not really much use in trying to understand what is going on in the structure. This is the equation for Von Mises Stress (in terms of principal stress - there are many other forms but you get the idea):

Von Mises Stress.png
 

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
Here are my calculations for the buckling of the 0.020" thick flange on the spar which has buckled in the test (3 pages from the textbook I refer to and one page of calculations below).

Agreement is pretty good with the buckling observed in the test (7242 psi vs 6900 psi). Note that buckling calculations always over estimate what you measure in a test and that is simply because real world structures are not perfect and things such as slight eccentricities in the structure will cause premature buckling. Also in this case the leading edge skin might have buckled twisting the flange slightly out of plane (not sure - cannot see the bottom of the skin in the photos). Anyway the calcs are within about 5%. I had a quick look at you calcs and they look like they are based on Bruhn. I will dig out Bruhn and have a look at what you have done in the spreadsheet.

Clearly in the test, the length of the flange is longer than the wavelength of this buckling mode and so it buckles into an S shape .... that is normal buckling behavior.

So the question we now face is, do we accept that this flange has buckled, will not carry more load, and press on with the loading ? I think the answer to this lies in what margin we have in the cap against buckling and whether we think this flange that has buckled is helping to stabilize the cap. Some analysis of the cap buckling is the next thing to investigate ......

Also, one other consideration is that if this is allowed to buckle, what does it do to the fabric if the wing is stressed to the point where it buckles in flight ? Any chance of ripping the fabric ?

IMG_8665.JPGIMG_8666.JPGIMG_8667.JPGIMG_8668.JPG
 
Last edited:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,787
Location
Saline Michigan
The problem I have with including any of the LE skin in the calculations, other than the narrow strip that is riveted to the flange of the spar, is that when the bending load in the skin gets to the root rib it has nowhere to go. So any bending load in the skin, in the area of the wing root, must be transferred from the skin into the spar and this implies that the bending stress in the LE skin at the root rib drops to zero. If this is the case, there is a fairly complex state of stress in that skin in the root area. Billski, I agree with your suggestion on how to treat the skin in the analysis if we are analyzing a section a few feet outboard of the root rib (and assuming there were enough leading edge ribs and/or spanwise stiffeners to maintain the shape of the D box skin on the tension side), but in the bay between the root rib and the first rib outboard, where the buckling is occurring, I think it safest to ignore any bending contribution of the D box skin.
Usually, I am the one pointing out that we generally have to funnel all of the bending and shear to the spar when we approach a connection point between wing panels. Ok, some airplanes rivet in a flange on the skins to connect them across a joint. Anyway, yes, the skin contribution to carrying all loads goes away as you approach the root end of the skins when they are free. While the spar alone is carrying bending and shear there, the skins do carry significant loads as we approach the connect point from the tip, and this is why many designs still beef up the skins near the connection. Common in composites is extra plies near the ends. In metal airplanes, stiffeners and ribs near the junctions are also common.

I have a suspicion that the buckle in the skin at the root might be due to some sort of warping of the D box because of the lack of restraint of the skin by the root rib but I need to think that through .... one step at a time. I also have some concerns about the strength of the root rib in general including the main section behind the spar but we will also get to that in time. Before that, we need to figure out the load path for the torsion load in the D box and how this is reacted into the fuselage.
I share the suspicion that the skins and ribs near discontinuities are highly stressed and may be difficult to analyze.

I think the next step is to sort out what to do about the flanges of the spar that are buckling and to make sure that there is enough column strength in that spar, otherwise more nose ribs or stiffer web stiffeners might be required to stabilize it.
Concurred (That means somebody else thinks so)

Billski
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,787
Location
Saline Michigan
Two things leap out at me.

First, the spanwise stiffeners shown in the nice FEA plots may be the wrong type of stiffening. I suspect that a somewhat beefy end rib and a closely spaced rib or two between the end and the next full rib stabilizes the skin and facilitate transfer of loads from the skins outboard to the spar at the root end of the wing panel. Being able to observe the direction of principal stresses in the skins near the root will be helpful here. Perhaps even a combination of an end rib and spanwise stiffners will significantly improve things. I also repeat my caution that if the wing is too soft in bending in the first place, it is likely to show up on the skins...

Second, the good agreement between elastic stability modeling and reality for the angle buckling is amazing. Dude, you are up to two bottles of good scotch owed to your analytical friend. Given how early the angle buckled, I suspect that significant stiffening is needed here. As discussed earlier, two paths exist that can be added to the current test article - one is to put a return flange on the angles and the other is to laminate additional material to the spar caps as you approach the root. Fussy? Yeah, but doable.

These two issues appear to interact. I would urge some iterating and classic calcs as well as that sexy FEA. Adding either a return or some more spar cap material is doable. Adding a false rib or three near the root plus a few spanwise stiffners might make big progress and be far easier to include for test. As usual, I favor the lightest scheme that works, because Weight Is The Enemy.

Billski
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
I suspect that significant stiffening is needed here.
Also, one other consideration is that if this is allowed to buckle, what does it do to the fabric if the wing is stressed to the point where it buckles in flight ? Any chance of ripping the fabric ?
I have no desire to continue loading without attempting a fix.

Some of my numbers look similar (.016x18 and .016x21, .036 x 42)) except my calculations are for a large span of .016 leading edge which I assume also caused the thicker .036 flange to bend. This assumption may be incorrect.

The worse bending was at the outboard attach of the drag spar (third rib) inboard for 6" . There was twisting of the 1st rib which caused the top portion LE just forward the rib to pull down. There probably also was twisting of the second rib,....I think it makes sense to calculate for the first rib but perhaps the moment at the 3rd rib is less, which would give a buckling load of even less.

PM if you need Bruhn 73 C7

Note revised spread sheet:
 

Attachments

Last edited:

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
I have a copy of Bruhn, I just ran out of time when writing the post above to look up the section you were refering to. I have a lot of structures books in my collection but that 1973 edition of Bruhn is the most used because of the example based approach that it uses. If someone teaching themselves stress analysis of aircraft structures is going to buy just one book, this is the one they should buy. It is nearly 50 years since it was published and they don't write'em like that any more. By the way, the 73 edition was not the first and the earlier edition had a chapter on the design of wooden structures (which is relevant to some homebuilders), but that edition is hard to find. Anyway I digress.

Do you have a photo from the test that showns the lower D box skin in the root area from the outside ?
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
Do you have a photo from the test that showns the lower D box skin in the root area from the outside ?
Best I have are at post 1.....the bare aluminum and reflections made it hard to see anything through the camera lens.

I can load it up again and see if I can get better pictures.
 
Last edited:

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,565
Location
NJ
The oil canning of the LE was extensive and from about the middle of the first and second panel from the radius of the point forward the plane of the end of the rib back 45deg to the bent flanges. They were large in the middle ellipsoid and small at the flange and radius. I think the drag spar and aft structure supported the first 2 panels to some extent with perhaps less strain of the spar until the middle of the second panel. Deflections are hard to measure because of the rubber wheels and deflection of the holding fixture. I did do FEA of the deflections but the whole structure and ribs were not in the model only the spar and LE so it probably is not accurate.
 
Last edited:

christos

Well-Known Member
Joined
Nov 9, 2018
Messages
52
Location
Greece
Thread is interesting but big also. I read it now and maybe I miss it. But I didn't see (in calculations) the moment of air load. Wing skin helps a lot in this load condition(it is the reason for +-45 wing skin fabric in composite structures). Last but not least, stress in wing spar will be upgraded because of torsional movement.

#simscale is using Code_ASTER,which it is a very good software, but I don't know the abilities of simscale's platform.
It is high recommended, when you are looking for buckling, to go with 3 dimensional finite element model. A good grid of elements is also necessary.

A good option is to use cae Linux, it includes all necessary softwares. Also Salome meca can run in Linux. And it is free to use.

In any case, a 3d element model with an excellent grid of elements is necessary.
 
Top