Clarification on what exactly the G force is here?

[theStruggler]

Hello,

I am trying to carry out FEM analysis on a steering knuckle. I have found some articles with typical braking and cornering loads to apply and they are all expressed in G-Force (here are two the articles I found:

http://iosrjournals.org/iosr-jmce/papers/ICAET-2014/me/volume-4/8.pdf?id=7622

and

http://www.altairatc.com/india/prev...optimization_and_finite_element_vitpune.pdf).

The braking force for instance in both is 1.5G. I would like to know what exaclty that means this case?

Is it 1.5 x the whole car's weight or is it 1.5 x the weight of the car supported by that wheel (1/4 of the total car weight)?

Thank you for your help.

 

[Nereth1]

Not an automotive engineer, but having followed some discussions here, it's almost certainly saying the whole car should be assumed to be able to decelerate at 1.5g and distribute the force to individual tyres as you see fit for your situation. E.g. if front ttyres have 3x the load of rear tyres, that needs to be appropriately included.

Which is fairly likely considering load transfer.

 

[BrianPetersen]

They are whatever the actual force is applied at that location when the car is cornering/braking/accelerating at that factor times the force of gravity, and don't forget that gravity (downward) remains in effect while cornering/braking/accelerating, and don't forget that cornering/braking/accelerating changes the corner loads on each corner of the vehicle.

Whatever the net vector sum of the forces that happen when cornering/braking/accelerating at that factor, is what the load is.

 

[GregLocock]

so if we go back to 3 2 1, an the static ground reaction at that wheel is 3000N, then the loads applied will be 9000 6000 and 3000 N.

make sure you apply them as separate load cases and then consider the various combinations when you add them up to find max stresses.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[theStruggler]

Thank you for your replies. So the G-force in this case will be whatever is supported by the wheel at static condition then.

@Greg I was thinking of applying all the forces at once (mainly braking, cornering and gravity) and if the part withstands that, then start cutting out material where not really needed. That is what the articles I found do. I did also find one that applied them as a load by load case though (

http://es.scribd.com/doc/99494468/C-Kalavrytinos-FEA-of-a-Steering-Knuckle).

Would you say it is best to go case by case and then combine the loads?

 

[BrianPetersen]

You had better make sure your parts are suitable for (let's say) X g vertical acceleration, and X g straight forward acceleration combined with static weight, and X g braking combined with static weight, and X g cornering combined with static weight, and if there is anything funny-looking that might look weak with a specific combination of any of the above, you had better analyze that, too. I am not going to speculate what "X" needs to be, in any of the above. People do drive into kerbs, over rough railroad crossings, and through rough washboard conditions now and again.

 

[GregLocock]

The stress at any point is a product of the forces applied times the sensitivity to each force. There is no guarantee that the sensitivity is positive.

As such the worst case stress at a particular point is not necessarily incurred by the combined loads in the positive direction.

This is amateur hour stuff foe FEA analysts, any paper you read suggesting otherwise can be discarded and ignored.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[theStruggler]

That makes a lot of sense. Thanks.

"I am not going to speculate what "X" needs to be, in any of the above" How would you say I should go about determining what X should be then? Like for braking and cornering I am using 1.5G as in the articles but for cases like bumps I am not exactly sure how to go about it.

 

[GregLocock]

Plenty of non aero circuit cars have been designed to 2g longitudinal 1g lateral and 3g bump (hence 3 2 1). You will get suspension failures under abuse conditions if you stick to those. If you are using steel rather than aluminium, and you design to yield at those loads, then the thing won't fall apart.

In my experience for road cars that are designed not to break those are low, the exact numbers we use are proprietary and I'm not going to talk about that. We also design our suspensions to fail in certain specific ways in response to the 60 kph square edge pothole test, which all OEMs have some equivalent to. We also have a running over the kerb test, and a smashing the car sideways into a kerb test (big lose on a corner). There's no simple static equivalent loads for those, they are architecture dependent.


here's a couple of other threads on this

http://www.eng-tips.com/viewthread.cfm?qid=300205

http://forums.autosport.com/topic/121354-kerb-strikes-and-wave-forms/

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[theStruggler]

Thank you very much Greg for your input and for the other threads you directed me to. It cleared up a lot for me.

I subjected my knuckle (aluminum 7075 not steel) to the individual 5-4-2 loading scenario and then combined the maximum stresses (square root of the squares). In each loading case, the weight supported by the wheel was always present. I got a stress value which was less than half of the ultimate and yield tensile strengths but greater than the fatigue strength. At this point, would you say I can assume my design is safe? Or would you recommend otherwise? I tried out a couple of combinations but got a maximum stress values I got were less than the one I got from combining the individual ones.

Thank you again!

 

[NormPeterson]

I imagine you could perform a cumulative fatigue damage assessment to estimate whether it will hold up for X amount of time. Choosing a representative variety of loading scenarios and estimating how many occurrences of each might be a bit tricky, though.


Norm

 

[GregLocock]

If it passes 5 4 2 then it probably will be ok. What elongation at failure does 7075 have?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[theStruggler]

It is about 11% at failure. I got all my properties values from here

http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA7075T6.

Something that concerns me about those values however is that next to them they have "AA" annotation and according to that page anything value with "AA" should not be used for design. Any thoughts?

 

[GregLocock]

Yes, 11% is too low for use on suspension parts. 15 is ok, 18 is better. For track use you'll get away with it if you aren't running the kerbs.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BUGGAR]

theStruggler,

The website notes that the values are from the Aluminum Association and are not for design. The Aluminum Association has design values and methods to use for aluminum for buildings and bridges. They also publish typical stress values which are typically noted as "not for design".

I have the Aluminum Association Design Manual, although it is primarily for building and bridge structures, it still has good data on aluminum.

Let me know if I can look up something for you (anybody else, too).

Bob

 

[theStruggler]

"Yes, 11% is too low for use on suspension parts. 15 is ok, 18 is better. For track use you'll get away with it if you aren't running the kerbs." Would it be possible for you to explain this a little to me? Also would you recommend using a different material in this case?

Bob I will be sure to take you up on the offer once I have a clearer idea of what we should/will use for the knuckle. Thank you very much for your willingness to help.

Thanks!

 

[GregLocock]

The more post yield elongation the more robust the part is.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[theStruggler]

Got it. So I will be right in concluding a different material will be more suitable then?

Hey Bob, will it be possible for you to look up the actual elongation at failure of aluminum 7075? If it is not asking too much, can you also look up the ultimate, yield and shear strengths and the modulus of elasticity for me?

Thanks!

 

[BUGGAR]

I looked at my Aluminum Association Design Manual and as I noted it generally covers aluminum for building and bridge design. It doesn't cover 7075.

ASTM Spec B 209 covers 7075 as well as all the other alloys - I would give you some values but it depends on the temper and there are several. I was able to download a '96 version of this spec. If you poke around you can find it too. Let me know if you have any problems. The Aluminum Association gets their design values from this for those alloys that it does cover.

Don't forget your safety factors. In the building industry, we use around 2/3 of the yield stress for design, but that's for buildings carrying our sons and daughters. A race car carrying daddy gets a lower safety factor.

Weldability may be a big problem with 7075

I like to work with 6061 T6 because it is a relatively "cheap" alloy. What everyone says about ductility (elongation) is important. I like to take a bar of 6061 and bend it in a vise until it breaks just to get a feel for how much it might bend before it snaps. Remember the song, "You picked a fine time to leave me loose wheel"? The higher aluminum tempers are quite brittle (from a guy who designs steel to bend radically, back and forth, before the earthquake finally snaps it.)

Bob

 

[theStruggler]

Thank you Bob for the ASTM reference. I was able to find a prior version too.

I have a question about the FEM analysis; I reran the individual loading cases changing where I apply the loadings and where I constrain the DOFs. Again, the max stress I got was less than the ultimate tensile and tensile yield strengths by a bit but this time it was slightly bigger than the shear strength. Should this be of concern or can I still safely conclude the part should work?

Thanks!