Wheel center FEA constraints and loads?

[Nihal]

We are planning to design our own wheel centers after using Keizer wheels last year. We were wondering how everyone constraints and applies loads to their design using a FEA program. We were thinking of simletaniously applying a cornering loads (applies bending to the spokes) and braking loads as a worst case scenario. I'm guessing I should also apply a vertical load from cornering/the weight of the car to the spokes as well.
Should I constrain the wheel by fixing the outer sruface/lip or the inner hub portion of the center?
Thanks

[Nihal]

We are planning to design our own wheel centers after using Keizer wheels last year. We were wondering how everyone constraints and applies loads to their design using a FEA program. We were thinking of simletaniously applying a cornering loads (applies bending to the spokes) and braking loads as a worst case scenario. I'm guessing I should also apply a vertical load from cornering/the weight of the car to the spokes as well.
Should I constrain the wheel by fixing the outer sruface/lip or the inner hub portion of the center?
Thanks

[Denny Trimble]

I benchmarked a few different load/restraint options when I started doing wheel center FEA two years ago. Try different scenarios, and use the most accurate one that runs in a reasonable amount of time. Before Cosmos/Works would do remote loads, I modeled a dummy wheel shell that had a stiff extension to the contact patch, and applied loads there (cornering, braking, vertical). And, I restrained a simplified model of the hub, and connected the wheel center to the hub.

With the current version of cosmos, I'd use remote loads to apply a contact patch load to the wheel center outer flange, and "connectors" to connect the lug bores of the hub and wheel center, instead of bonded contact.

You should look at various load cases; I concluded that cornering was the worst case by far, in terms of stress and deflection.

One big question area, especially if you want to look at analyzing the wheel shells, is how does the tire load the rim, exactly?

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Denny Trimble:
One big question area, especially if you want to look at analyzing the wheel shells, is how does the tire load the rim, exactly? </div></BLOCKQUOTE>

Exactly? Good question. Maybe ask the guys who are doing the tyre testing? I'm curious too.

My guess is that the tyreprint can only load the wheel shell via the sidewalls and beadwire. The air pressure will always be fairly uniform around the shell, hence only a uniform compressive load there (whether the tyre is off the ground or under high vertical/braking/cornering loads).

My guess is that a vertical load (Fz) acts from tyreprint, through sidewalls and air pressure, to suspend the wheel from the beadwire. So mainly radial (towards hub) loads at the beads from about 5 to 7 o'clock, with lesser radial loads to 4 and 8 o'clock.

Acceleration/braking loads (Fx) will be via friction at the beads, so circumferencial plus some radial loads. If forwards is to the right then braking (plus the necessary vertical Fz force) will have most of the bead loads from about 3 to 7 o'clock.

Cornering (Fy) distorts the tread belt sideways, so gives greater lateral beadloads on one bead "lip" and lesser lateral loads on the other bead lip (normally both beads are pushed outwards by the air pressure). In general, a crossply tyre will have the lateral bead loads concentrated towards the bottom of the wheel, while a radial will have more uniform lateral loads from bottom to top of the wheel. However, the radial (usually!) has greater lateral movement of its tyreprint (and the whole tread belt), so the Fz loads move further sideways than for the crossply...

Anyway, that's my guess. Corrections welcome http://fsae.com/groupee_common/emoticons/icon_smile.gif.

Z

[Nihal]

Ok Denny that makes sense. I thought about it some and it seems that if I constrain it at the the hub face it would be realistic as in cornering the hub face takes the load from the wheel center to the upright.
In FEMPro/Algor I'm guessing I could make a shell object or a beam structure that linked up to the outer wheel cetner face and applied the loads there. I don't think Algor does remote loads (all the moments we try to use come out as 0 for some reason). I'm wondering if I could just apply the force to the outer wheel center face, but I am not sure how to make the load properly spread over the face. The amount of load would decrease further from the point of the tire contact patch force.
Does Solidworks give you Cosmos when they give you the free student version they give to all FSAE teams?

[Denny Trimble]

Yes, the "Student Edition with Cosmos" is what you want. It retails for $200, and the "Student Edition" (without cosmos) is about $90.

[TG]

You can get SolidWorks 2005-2006 Student Edition with COSMOS here (http://www.studica.com/SolidWorks/#2005-2006%20Student%20Editions) for $150 including shipping.

[Nihal]

Denny: Would it make much of a difference to just restrain the wheel at the hub face instead of making a hub model and making my part an assembly? I'm guessing you either fix the part or allow only rotation about the hub axis.

Z: that all seems to make sense. Why do you say that a radial tire will concentrate the forces at both the top and bottom of the beadwire? Is it because the tire will kind of pivot and push against opposite sides of the beadwire at top and bottom? Or is it because the chord construction and tire design etc.?

I got the CAD of the kiezer wheel center + shell off the Keizer website but I can't import it to Algor for some reason. It can be converted to a 3D solidworks model, but when i try to put it in Algor it says "no parts were imported" or it only imports a portion like a bolt, or a piece of the shell. Anyone know why it isn't working? It would be nice to have the entire tire shell for FEA purposes.

What weights and stresses are most people seeing with their wheel center designs. I am a little confused as the Keizer wheel center that I FEAed came out as a rather low stress. I applied a 800lb cornering load to the outer face of the outer rim of the cetner and fixed the hub face of the wheel. I got a max of around 5200psi. People have told me that the Keizer's flex alot and eventually fatigue, but this is far below the fatigue strength of Aluminum. What am I doing wrong? I should try it with a dummy wheel shell next time I guess.

The two design's I cam up with using 6061 were around 1.8-2.51lbs but they were horribly overstressed, but I don't trust the way I have set up my FEA. Any have any input to this?

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Nihal:
Z: Why do you say that a radial tire will concentrate the forces at both the top and bottom of the beadwire? </div></BLOCKQUOTE>

Nihal, Just my observation of radials vs crossplies. Watching the two of them during cornering, it seems that the whole tread of the radial moves sideways, relative to the wheel, so I guess that bead side forces are pretty high at the top and middle of the wheel, as well as at the bottom (where the bead side forces would be greatest). With crossplies it seems as if the bottom of the tyre is most distorted, with less sideways movement at the top. In both cases the forces are in the same direction, top and bottom.

Again, that's just my observation. It would be good to watch the Calspan tyre testing that is soon to happen. Or maybe just some pics of the tyres during testing, hey, Denny???

Z

[Denny Trimble]

The registered teams will get DVD's with data and video footage. But, they won't be allowed to share or publish the data on the internet. They will be allowed to publish it in academic papers, with proper citation as spelled out in the agreement.

Here are some interesting slip angle visualizations:

http://students.washington.edu/dennyt/fsae/RWLG1.jpg
http://students.washington.edu/dennyt/fsae/RWLG21.jpg

[Denny Trimble]

Oh yeah, and this is from an Ohio State video (cut and slowed down by myself):
http://students.washington.edu/dennyt/fsae/tireclip.avi

[Jarrod]

Something that may also be relevant is that this paper (2004-01-3527 : Track Surface Effects on Race Tire Performance) suggests that the sidewall/carcass deflections shown in the video may be the casue of load sensitivity in tyres