prkiller

This post is to address these debates posted here!
https://honda-tech.com/zerothread?id=1385475
https://honda-tech.com/zero...age=1

There has been much speculation about which brace is stronger. I asked for the help of a fellow HT member to run an FEA analysis on both braces. We applied 500lbs of lateral load (generated by LCA) and 700lbs of twisting load (generated by the swaybar).

How to read

Blue color means lower stress and Red means higher stress.

Yield strength is set to 8.000e+03 psi for a 6061 T6 material.

The colors on the different pictures means the general amount of stress on that particular area. You must look at the numbers (psi) for reference to pin-point the highest area of stress. The higher the value the higher the stress. Red representing the highest value!


Lateral Stress




Twisting Stress




Twisting Displacement





I hope this simulation can end the debate about which brace is stronger! Will they both do the job, sure looks like it from this sim. Is one stronger than the other, again, sure looks like it from this sim. Really, this sim was done so we could see which was "actually" stronger. Now that everyone can see these results, hopefully everyone will get off the whole "I beam design is the greatest thing since sliced bread" act!
You can see that in fact the I beam design is VERY strong, but the ends will likely break off before the I beam can do any real work! This in no way implies the brace is bad, just stating an obvious fact that is supported by the sim.
I for one am happy with my ASR and glad that the numbers show what I thought was the case from the beginning!



Modified by prkiller at 8:52 AM 10/11/2005

slammed_93_hatch

thanks for having this done, but im confused on the how to read the stuff.

the color shows were the highest level of stress is placed on both right? What does that number show then? is it the most stress that will take befor bracking or is it how much stress is generated in that area?

prkiller

Well, from what I understand (because I am no professional engineer), it is the amount of stress that will be exerted on that particular point. Pay close attention to the yield strength because that number represents the point at which the piece will fail. Anytime the number is higher than the yield strength on the piece, it will fail at that point.

These are all calculations based on the original numbers (500lb, 700lb), and as such, the numbers generaged are in relation to those. If you change the original numbers, your final calculations change as well, but the relationship between the ASR and comptech brace in relation to strength does not change!

Eyal 951

its nice to see testing finally done, and to finally put the I beam argument to rest. Those were results I expected to see, and I'm glad somebody did it. Charleston knows what he is doing, and ASR is an up and coming star for aftermarket Honda parts.

nonsense

That's awesome guys. That is exactly what I was looking for. Thanks for doing the tests.

BTW what software was used for the FEA? -

PIC Performance

Looks like Cosmos, thats what we use for our FEA analysis too (only in Chinese )
Hey PRkiller, quick question since you have used Cosmos: do you know if the software can also perform fatigue analysis, and if so, have you used it on this sim?

nonsense

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by PIC Performance &raquo;</TD></TR><TR><TD CLASS="quote">Looks like Cosmos, thats what we use for our FEA analysis too (only in Chinese )
Hey PRkiller, quick question since you have used Cosmos: do you know if the software can also perform fatigue analysis, and if so, have you used it on this sim?</TD></TR></TABLE>

Looks like the pro and adv. pro can do cyclic fatigue analysis.

http://www.solidworks.com/page....html
their FloWorks looks cool too

ecahatch

vwong

The models were done in SolidWorks. The FEA anaylses were done in COSMO Express. Good little tool when you need to do some FEA on some parts.

cxSHOE

awesome work... i wonder if the results are the same for the dc2 version of both?

prkiller

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by cxSHOE &raquo;</TD></TR><TR><TD CLASS="quote">awesome work... i wonder if the results are the same for the dc2 version of both?

</TD></TR></TABLE>

I would very much think so!

prkiller

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by PIC Performance &raquo;</TD></TR><TR><TD CLASS="quote">Looks like Cosmos, thats what we use for our FEA analysis too (only in Chinese )
Hey PRkiller, quick question since you have used Cosmos: do you know if the software can also perform fatigue analysis, and if so, have you used it on this sim?</TD></TR></TABLE>

This is your man! He knows all! Well, maybe not, but he's a pretty smart guy!

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by vwong &raquo;</TD></TR><TR><TD CLASS="quote">The models were done in SolidWorks. The FEA anaylses were done in COSMO Express. Good little tool when you need to do some FEA on some parts. </TD></TR></TABLE>

LowTEC-Derbo

thanks for settling the debate.


Made my decision on which one to get alot easier.

.water

fak iam a noob. soo computech one is lil better?

Eyal 951

no, ASR is... look at the numbers.

Type V

SWEETNESS!!! SolidWorks used to its FINEST. very happy to see it done!

did you actually have both items to measure "at hand" and thus model into SolidWorks?

or were the dimensions used to model the two parts based off actual dimensions provided by both ASR and Comptech? &lt;--- which i doubt?

either way, i guess even if you "eye-balled" the dimensions as best to fit/match onto the actual length of the subframe and mounting hole locations, it would still suffice based on overal "shape"!

If I knew this 4 years ago... I would have went ASR... unfortunately I went with 'typical' reinforcement plates prior to the availability of Beaks... and thus tore my stock frame ---&gt; end up welding in the CTR frame!

very good news for everyone now! GOOD JOB!

vwong

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Type V &raquo;</TD></TR><TR><TD CLASS="quote">did you actually have both items to measure "at hand" and thus model into SolidWorks?

or were the dimensions used to model the two parts based off actual dimensions provided by both ASR and Comptech? &lt;--- which i doubt?
</TD></TR></TABLE>

I had the actual Comptech piece in hand at the time I modeled the part into SolidWorks. And I was able to obtain a copy of the drawing (with dimensions) for the ASR piece. So these models are pretty accurate in representation of the real pieces. BTW, both of them are the EK braces. I don't have any DC braces to work with.

Tyson

these numbers looked pretty funny to me. so i asked my friend to take a look and comment. he has some interesting things to say...

(he's a stress analyst at northrop grumman that builds satellites and got his masters in composite structures from stanford)

-------

Here’s some questions I’d raise to the analysts who did the run.



1.) Why did they use 8000psi as a yielding stress (which is the yield strength of nylon)? 6061-T6 Al has a 40000psi yield stress.



2.) How did they constrain the part (ie how is it bolted on)? How were the forces applied? It seems as if the strut bar (taking the comptech plate as an example) support brackets bolt on to the plate and that plate bolts on by two holes to the car chassis? If so, I would put loads on all 4 smaller holes and for the 2 bigger holes, I’d constrain them like a fixed end so that they don’t rotate or translate. If this is the case, I’d expect to see a lot of stresses concentrated at those bolt holes that I constrained. Some of the pics of the stress countours have minimum stress contours which in my mind raises a red flag. I’d really like to know how they did that because how constraints and loads are placed and how they are created are oh so important for getting accurate stress results.



3.) The displacement numbers are trash. 1294mm (1.3meters for the comptech plate) U resultant disp for Aluminum of that thickness under 500lb total lateral and 700 total loads caused by moments? If I read that plot right, then you have to be kidding… Also they say it is a twisting stress which implies a torque and I think the contributions to stress by torque is FAR FAR less than the contributions to stresses by lateral and moment loads.



4.) Also they also just mention the stress and how low it is. They totally forget to mention the fact that it should be a weight concern as well for that kind of part. I don’t find it useful that they say that basically the part is less stressed, and therefore, it is better. I can beef up the part to hell, replace the material with stainless steel or titanium and have better stress results. But a ratio of max stress to weight will be VERY helpful for me.



5.) Lastly, they forget that the von-mises stress is a failure criterion. Anything less than that 8000psi number is of little to no concern. Anything above and that is an issue. (but they idiotically use the value of nylon instead of Al so of course, everything is pretty much going to fail.) One of my books lists a Factor of safety using the Von mises stress as (yield stress/mises stress). That is what ultimately gets told as the result.

That’s all I can really find for now, but it was interesting looking at it. 1.2m displacement. Haha that is hilarious. I hope some of this makes sense….



Tim

Type V

NICE - very happy to see the use of SW!!! I just learned it at our local tech institute - had a contract job after co-op, but that ended, and now employers are STILL saying i'm too inexperienced - bummers.

incidently, your screen name is actually my First Initial and Last Name!

vtecvoodoo

Tyson the suspension/engineering **** strikes again

Tyson

not my comments. just my friend and coworker's. but the whole point of a non constrained body was obvious to me. its really just pictures with pretty colors...

vtecvoodoo

I sent it to an engineer/buddy at Comptech. We'll see what he thinks

vwong

This was the default figure from COSMO Express. I double checked the yield stress for 6061-T6 AL my book, it is 35,000 psi on tension yield stress and 20,000 psi on shear yield stress.

Both braces were constrained at the LCA bolt and the middle hole on the far end. I don't know about the full version of COSMO, but COSMO Express won't allow me to apply loads on the same surface where I applied constrains. I tried applying constrains differently, but the results were even whacky.

For the sake of comparison, if the constrains and loads are applied the same, what is the issue? Ever heard of "worst case scenerio"? Applying constrains to only one end ressembles this.

Well, if the braces were constrained to only one side, of course the displacements are going to be large. This is just to compare how flexible each brace is. Oh, and in the displacement analysis, only the 700 lbs load was applied and it was by moment.

I agree to this. But again, this is a comparison of 2 already available parts. If you're designing a part from scratch, and weight is a big factor. I can totally understand that.

This will change if the yield stress value is updated to 35,000 psi.

GChambers

I get the feeling we are just splitting hairs here. Am I right?

Also, I gather from this thread that neither is going to break under anything but the most extreme circumstances. Am I also correct about this?

Eyal 951

well, I've heard of people saying that they can flex the comptech brace by hand. The i beam portion my be strong, but its ony as strong as its weakes point which is the sides, that are thin, i beam less, and flexable.
I can't flex the asr brace by hand in any direction.