complexx

bruinbro,

I would really like to thank you for all the time and effort you have put in so far. Its really beyond the call of duty for an online forum.

The table will be quite useful for the members to go back to again and again. It is a nice resource to have.

One thing you should add to the table is the input load that produced the results so that anyone who examines it will understand that it's not a typical load for a side impact. We don't want anyone to think that in a substantial impact the cage will deform less than an inch

-Frank

ed devinney

Ahhh, punch cards. I'm not that old, but I did run my share of 'em as a kid.

This is good stuff, thanks very much to all for collaborating.

bruinbro

Are we done?

complexx

Oh, I was just saying thanks. I'm actually looking forward to seeing some more detailed stuff if we can get the data from someone with a cage that can map out some points, etc. This is my first time dealing with the 944 so I'm not totally sure as to the best locations for the mounting points of the rest of the cage yet. Hopefully someone can provide the data you need to model a more detailed structure.

I am in no way trying to end this thread

camber799

This may be true but only if you have a race logged in yourlog book with the cage configuration that is acceptable at the time of the event you attended.

If you have never raced with that sanctioning body you will be subject to the current rules.

Rulebooks do change so theoretically you could find that the sanctioning body may not grandfather your cage. It's unlikely but it could happen.

Generally I would look at SCCA requirements as the baseline and then look at the rules of the sanctioning body you are planning to run with. If one of them has a requirement for a heavier material I'd do it. You should then be legal in both.

gbaker

That's nothing. I have a Pickett slide rule around somewhere.

You stand in line at the punchcard machine, then stand in line at the card reader, then try to get one of the machine operators to estimate when your "job" will run so you can decide whether to come back the next day or grab a quick bite to eat and return for your printout. You usually had to wait for your printout (because something broke somewhere) only to discover that your job did not compile because of unbalanced parentheses.

And as Bro pointed out, the best approach was to do all this at 2 AM. It was about as much fun as chewing on broken glass.

Geo

Damn. I used to have a slide rule. I'm probably one of a handlful of people my age who has ever used one. By the time I hit high school you could buy a gazillion function calculator for $20.

Speaking of punch cards.....

About 15 years ago I started a new job and my boss was walking me by the IT department. He said "And this is where the ladies punch the 80 column cards." I started laughing. He asked me what was so funny. Then I saw he was telling me the truth and I was wondering what I had gotten myself into. Up until 10 years ago they were still punching 80 column cards.

95m3racer

Cool stuff, unfortunately, sorry to burst everyone's bubble's, but the FEA does not mean much.

Crash Analysis is one of the toughest things to analyze, and when you are considering a production based unibody, with a couple tubes welded in...its pretty damn complex.

While the analysis of the door bars might seem conclusive (to everyone but the person whodid the FEA analysis) it does not mean anything in real life.

Look at what REAL nascar bars are, they go upto your shoulder, and consist of 4-5 rows of tubing in each direction. The "nascar" bars in club racing are merely smaller versions that do the job, but are "nascar type" at the most. Couple not the best welding, with improper base plates, large gaps being filled with weldfiller, have fun accurately modelling that in Nastran

As far as what protects best, almost every factory race car, from BMW to porsche to ASton Martin all use X bracing, some with gussets, some without, some with extra bracing to Apillar, some with ridiculous amounts of extra bracing.

As Bro, the resident FEA analyst, should know, FEA tests will look accurate to everyone but the guy who made the model, b/c he knows what he assumed, and what's not accurate as far as all the parameters being used.

As for pounch cards...the best use we found for them was dumping them down the middle of a 10 floor high stairwell They make some cool sounds when falling and crashing on the floor. (Complexx had a part in that haha)

Aston DBR9: Notice how the xbrace is made...
http://www.dieselstation.com/archive...R9/index.shtml

BMW Factory E46:



Another example of a different style X brace and A pillar bracing.



Factory renault tube frame megane trophy cars. Note: They also use a thick barrier of "energy management" between the xbrace and the door, similar to what BMW does with WTCC cars.

M758

I don't believe the FEA is a complete waste. It does not tell the entire story, but by doing a simple estimate like this you can gain some insights. I don't thing that the numbers thenselves mean a whole lot, but compare one to another and trends can emerge. What seems to happen is the rather basic arch shape of the nascar or pyramid bars tend to support a little more load than just a flat plane X. The theory before was that they would be too flexible. Seems now that it may not be the case. I do however think it seems to show that the pyramid is strong relative to its weight. This seems to support the idea that the NASCAR bars are good, but heavy solution.

95m3racer

The Pyramid is strong if impacted on the peak, and that the force is perpenciudlar. Also assuming a/b pillar is not moving, and all welds hold. Also assuming the chassis is not hit and deforming the mounts which hold these tubes. Also assuming the impact is at this one force/speed, and not changing due to other components of the car. There are literally hundreds of variables in this one area of the car alone.

The "theory" behind what will happen is what any structural/mechanical engineer will already know, FEA will not show anything past that, until you get to more advanced models that are of structures that cannot have a general theory applied.

Its like building a bridge...while some general ideas are basic and understandable/predictable, actually knowing what will happen needs a LOT of analysis and help from computers...but having proper assumptions made by humans, and good data from experimental tests is key to any successful structure/system.

However, as an engineer, I love to see this stuff on the forums, compared to some forums that post pictures of cages from some hick in the woods, and say, "oh yea dem dat dear is a nice *** safety bar."

I'm still working on my semi-tube frame design for this e46 m3 project i'm starting, let me get some more cad done and i'll post the ideas. Here is an older design i was working on, but its evolved since then. This was mostly a model to get a good a pillar design, the rest of the subframe reinforcement, rear end integreation, and front end framing was not included. I've now gone to a main a pillar support bar that is seperate of the X brace, like aston. Uploaded a quick fix to show that idea.

Anyone else have good example of "different" cage designs? Post em up!

Geo

I agree with this. It shows solid relative data. It certainly dispells some long held assumptions on my part that had (I thought) strong engineering basis. The absolute numbers are less important than the relative numbers.

This has me rethinking my door braces. I'm getting very very close to starting paint prep, but that means I have a small window of opportunity to redo the set-up.

When I get home I have an idea to salvage my current door structure in terms of relative strength with little additional mass. However, I'm tempted to cut the X out and install NASTYCAR bars. The relative difference between the pyramid and the NASTYCAR bars is of small magnitude and the increased space afforded by the NASTYCAR bars make them suddenly VERY enticing.

complexx

Are you a relatively large person? I was also wondering about how invasive the X brace is in the 944. Do you experience any interference?

bruinbro

Iguess one of the things I should do is estimate the relative weight differences. I'm somewhat hesitant as I don't know if the geometry is really relevant. Anybody care to chime in?

Hey George, where's the cage dimensions?

Bro

Geo

Um, yes, I'm relatively large. Currently 234.5 lbs (as of Monday) after losing 21 lbs. However, by the time my car is finished I expect to be 175 lbs or less.

At my former weight of 255-260 in my Sparco Evo2 I didn't feel encroached upon by the X, however, when you sit in a car with NASCAR bars and they are seemingly a million miles away you appreciate the advantages.

I wasn't worried about the X, however since the NASTYCAR bars are so close to the pyramid in performance and with the advantages of space, they are very very tempting.

Geo

Indeed. Added mass may very well affect the performance.

I should get by the shop tomorrow so I'll get you the dimensions then.