Geo

Most interesting. I may have to break out a sawzall before I paint my car.

kurt M

opinion to follow... Looks to me like Pyramid did slightly better for total deflection than the rest but the NASCAR bars start further out from you. total movment is small ovall in all 3 cases but so is the energy. If you have the room and don't care about weight or triangulation of the door hole -NASCAR. Care about weight some and care about triangulation of door hole- Pyramid. I see no reason for flat X other than you do not have room to make a pyramid. Gusseting is still out until we see a demo with double wall tacos. Single thickness gusset does little if anything when pushed from the side as it has no geometric strength in that direction. My bet is that a double wall gusset will only move the distortion out away from the center of the X. This might be the thing to work at. Use tacos in a way to firm up the driver’s area of the X.

Thanks again for running the models!

M758

I believe a gusset in the X part does nothing other than reduce the stress level at teh weld joint. This will not really change displacment much in model like this, but the gusset allows alot more load before a weld failure. Plust it provides a bit more intrusion protection since the X is at its smallest point the.

Now I think the results maybe different if you put a torsional loading on a cage with these door bars. Of course the torsional load is not crash load, but a cornering load. You really need the rest of the cage however to figure that properly.

bruinbro

Please look a the NASCAR analysis again. I inserted the wrong picture for the narrow loading scenario. The correct one is in there now.

Thank you, we will now continue with our regular programming.

Bro

bruinbro

How would you like the summary? Little pictures? A table of max deflections? A post with all the pictures?

Let me know.

Bro

gbaker

I love this stuff...

BTW Bro, nice work. What package are you using (too lazy to check prior pages, sorry.)

bruinbro

It's easy for me to re-run the model with gussets that are a different thickness which would be the same or stiffer than a taco. Changing the gusset/taco is a bit more work, so if that is what you have in mind I need a notional size and location.

Bro

M758

Just a table. The key is to make sure you are comparing the same load cases to keep things fair.

bruinbro

I'm using I-Deas FEM and Model Solution. The pain is having to screenshot the images to PowerPoint, add the notes and arrows and stuff, save that as a .jpg, upload that to my picture server and then post it here. Must be a way of automating all that....

Thanks for the kudos.

Bro

gbaker

Got it. I am sufficiently old and grey (to which Geo can attest) that I once ran FEM on NASTRAN via punchcards.

I can't help there, although post processing to MathCAD/Axium graphics might be an option.

Well deserved.

bruinbro

How's this?



Bro

M758

Bro NICE...

Hmm Looks like I might need a little pyramid action!

Hmm I guess that is why new cup cars and RSR's have these now!

bruinbro

BTDT. Not NASTRAN, but programming in original FORTRAN and COBOL. Punchcard machines were cool.

Bro

complexx

You guys must be OOOLLLDDD. I had a few professors in college that used to talk about the good ol' days of standing on line for an hour to use their punch cards, realizing they had a mistake and having to wait on line again and again and again.

bruinbro

Yup. Got to do that a lot in college. You learned to submit your deck at 2:00 AM so you didn't have the hordes in front of you. Did kill lots od rainforests, though, as a simple error would result in reams of fold paper printout.

This



Plus this



Equaled this



Bro