View Full Version : Finite Element Analysis for Cobra/GT brackets

03-27-2007, 04:31 PM
I just got the results back this weekend for the brake brackets I have
designed from the FEA engineer. The Finite Element Analysis (FEA)
was completed on the SolidWorks 2005 platform.

[u]Setup Data :</u>

[u]Using projected performance data :</u>
60mph to 0mph in 135ft.

[u]And equations based on projected performance data :</u>
vfinal=vinitial+a*t (vfinal=0)
d=dinitial+v*t+0.5*a*t^2 (dinitial=0)

Creates a calculated a=29ft/sec^2 deceleration which is about 0.9gís

Round up to 1g for a safety buffer.

3500 lbs @ 1g braking = 3500 lbs braking force.

Front tires produce about 70% of the braking force:
3500 lbs * 0.70 = 2450 lbs

Force per front tire: Ftire = 2450 lbs / 2 = 1225 lbs

[u]Force applied to caliper:</u>
Fcaliper=Ftire*(Rtire/Rcaliper)=1225lbs*(13in/6in)=2650 lbs

This load will be applied evenly to the two caliper bolt holes.

[u]Results :</u>

The peak Stress Results show 12,000 PSI at the top mounting bolt
and is considerably lower than the yield strength of the material
(the 6061-T6 is 35,000-45,000 PSI depending where you look). This
is a 'worst case', in reality the stress would not be quite as
concentrated as is shown in the model, so you would only see peak
stresses around 10,000 PSI.

The Displacement Results show we get around three thousandths of
an inch of deflection at the front caliper bolt and around one
thousandth at the rear bolt. Since in reality the steel caliper
bridge ties these two together the caliper (and bracket) should
experience a deflection magnitude somewhere between these values.


[u]Clarification on GT bracket vs Cobra bracket :</u>

Six Grade 8 bolts are used in the middle (Stude used only 3) & Grade 10.9
bolts for the caliper (same as Ford). The GT bracket is what I tested, it
would be the "weaker" of the two. Technically that bracket wouldnt see
the level of stress that was found, unless it was used with a Cobra caliper.
The GT bracket only uses a 11 inch rotor, which would reduce the leverage
the caliper has. I figured lets see the worst case, and I will know it is that
much better off. If you notice the twisting on the bracket and the deflection
would both be less if the caliper bolt holes were better tied together. That
is not a problem since the Cobra bracket doesnt have that half moon cutout
in between the caliper mounting bolt holes. That will stiffen the bracket. If
the later 2 piston GT calipers are used for the GT size brakes, the "Cobra"
bracket can still be used, since the only reason the half moon is there is to
clear the early single piston GT caliper.

[u]Further reading:</u>


"The ability to model a structural system in 3D can provide a powerful and accurate analysis of almost any structure. 3D models in general, can be produced using a range of common computer aided design (CAD) packages. Models have the tendency to range largely in both complexity and in file format depending on 3D model creation software, and the complexity of the modelís geometry. FEA is a growing industry in product design, analysis and development in engineering. The trend of utilizing FEA as an engineering tool is growing rapidly."


'63 Avanti, zinc plated drilled & slotted 03 Mustang Cobra 13" front disc/98 GT rear brakes, 03 Cobra 17" wheels, GM alt, 97 Z28 leather seats, soon: 97 Z28 T-56 6-spd, Ported heads w/SST full flow valves, 'R3' 276 cam, Edelbrock AFB Carb, GM HEI distributor, 8.8mm plug wires

03-27-2007, 09:40 PM
Never used the SolidWorks finite element platform. Looks like a cool set up. How many nodes can you go up to for modeling purposes? Our professor at UT-Knoxville gave us a student version of Adina. All the features of the full blown version but only 900 nodes for modeling. Sometimes had to be resourceful to reduce the coarseness of the mesh and improve the accuracy of the stress-strain-deflection answers. Unfortunately my license on even that has lapsed.

The brake brackets are structural aluminum?

03-27-2007, 09:42 PM
FEA is great stuff, very useful for exactly what you are doing. I haven't seen anyone else do this with Studebaker parts. Good job!

Maybe the on-line group can assist in an additional part of this kind of design: Failure Modes Effects and Analysis (FMEA). In this part of the design, we address what can fail and how bad can it get? This is usually based on only one thing going wrong, not an improbable chain of events. The stresses get recalculated based on a possible failure mode.

I'll submit leaving out a bolt or having one loose. Also, are there conditions that can generate large shock loads? - like sliding into a curb with the brakes locked. How about a spin to generate large side loads?

What else might stress these parts?

[img=left]http://www.studegarage.com/images/gary_ash_m5_sm.jpg[/img=left] Gary Ash
Dartmouth, Mass.
'48 M5
'65 Wagonaire Commander
'63 Wagonaire Standard
web site at http://www.studegarage.com

03-27-2007, 10:08 PM
Could get my sister who teaches 6th grade science to buy a 24 month license copy of SolidWorks for $80, do all the parts analysis I want and save $4915. Now that appeals to my Studebaker cheap gene.


03-28-2007, 12:10 AM
Honestly, I dont know much about SolidWorks, thats why I let someone
else do the analysis. As far as failure modes, the stresses you are
talking about (loose bolt) would cause the same failure as the part
already on your Stude - bolt failure or "hogged out" holes. This is
true of the backing plate on your drums, or the little disc bracket.

The other possibilities you mention wouldnt really effect this part,
a hit against a curb would cause the pads to slip - taking load to a
zero condition, the only way a side load could be generated, is if
the spindle broke causing the rotor to hit the caliper. At that point
you have "other" things to worry about - that brake is now useless in
that the wheel is gone.

The only time that the bracket will see max load is right at the point
the tire loses traction. Locking the brakes up causes stress to drop.
Stress is always radial, which is the parts strongest direction. The
graded bolts take up the transfer load to the spindle flange which is
torqued in place spreading the load out.

Pretty much everyone from the CNC machinist to engineers I have spoke
with have told me the part is way thicker than it needs to be. Thats
just fine with me, just following the Studebaker tradition of making
parts overkill.;)