New to FEA, designing a product. Are my results valid? [Solidworks] 2

[eclipseband]

I am designing a plastic skateboard part in Solidworks and could use some help determining whether my testing is valid.

Link to photo album:

http://imgur.com/a/1FAY8

I am trying to determine whether the board is thick enough. My rated load is 300 lbs since people will be jumping around on them though there not too many overweight skateboarders. The material properties were inputted for flexural modulus and strength (rather than tensile) as I saw those values are better to test plastics loaded in bending.

The test was set up with the load placed on a face that is around the middle. The boards and its trucks are bonded using a contact set since adding bold contacts made the program go haywire with large deformations and each failed test took an hour. The fixtures are on the inside edge of each of the trucks that way they could bow out as if they had wheels.

My questions/concerns are:

1) Did I set this up correctly? Should I make some makeshift wheels? If so does anyone have any good resources/tutorials on how to set up rolling supports?

2) I have a minimum FOS of 0.86 though this is right near the trucks where it is supported. Should I be concerned with this? The FOS in the middle of the board is around 1.5 which is great.

With these kind of boards it is desirable to have some flex to it in order to carve and corner so I would not want to make this too rigid. I am looking for a sweet spot but these stress concentrations are throwing me off.

Thank you for your help!

 

[btrueblood]

No, your results are not valid. The trucks should not be completely fixed (the way you have them) - they should be able to rotate as the board bends. Think of a simply supported beam, one end is able to rotate but not move horizontally or vertically, the other end is able to rotate and move horizontally but not vertically.

 

[GregLocock]

300 lb *0.86 as a design load is way off any acceptable FoS.

Cheers

Greg Locock


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

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

 

[eclipseband]

Thanks for the tip.

The minumum FoS is far from acceptable though should I be concerned if it is only such a small area? Is it relatable to when using bolts there is a stress concentration in the surrounding hole?

I would really appreciate some insight on this.

 

[GregLocock]

No, FEA in general is not very good at the little red bits. One approach is to extract the general level of stress in that region and then use a hand calc with stress concentration factors to check details.

Another is to ignore the little red bits if they are small.

Neither is totally satisfactory.

FoS is better termed factor of ignorance, in your case you really don't know the loads. But for many applications a FoS of 10 would be sensible for a man rated device.

Cheers

Greg Locock


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

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

 

[rickfischer51]

Your boundary conditions are hosed as btrueblood pointed out. Your stress patterns will totally change with the correct BC's, so fix that before you worry about the 0.86 FOS. Keep in mind that your load and results are based entirely on static conditions, but a skateboard is anything but static. You need to consider impact and fatigue, so a FoS of 1.3 is not very great. Also, you say the skateboard is made of "plastic." That's like describing your wife as female. There's a big difference between Rosie O'donnell and Angelina Jolie. What plastic, and how is it made? Is it injection molded? Is it glass reinforced? Your red bits at the trucks are at an area that may see creep because of fastener loads, knit lines at molded in holes, bosses or inserts, etc. so high stress/low FoS there is a concern. Is your material notch sensitive? You should probably get Structural Analysis of Thermoplastic Components by Trantina and Nimmer and give it a good read.

You've loaded the part in the center, and while that certainly results in an extreme loading, I don't think I've seen many people on a skate board on one foot, or with booth feet planted in the middle. Seems like the feet should be right over the trucks. You should be looking at multiple load cases. Go to a park and see how kids are actually using them.

Rick Fischer
Principal Engineer
Argonne National Laboratory

 

[eclipseband]

Thanks for the advice Rick. That book looks real good. I'm going to check that out for sure.

You're right. Doing this test is somewhat moot since the loading is dynamic and will be centered on the trucks though this is the best option I have. With my limited resources I can't make a real protoype.

I am using a natural fiber reinforced plastic for plastic injection which has some great properties. It behaves like a glass fiber reinforced material. My design has centered a lot around what others have done. Knowing their material (ABS) and their dimensions I think my design should work. As it stands now the board is 0.50" thick. The izod/charpy test values are higher than what's currently being used so to my understanding notch sensitivity and impact should be alright. Only concern with the material would be the elongation before break is 3%.

For reference I have attached part of my material data sheet.

 

[rickfischer51]

First, that is not a very ductile material. The rule of thumb is anything under 5% is considered brittle. The "red bits" mentioned above can sometimes be overlooked if they are small and the material is ductile because local yielding will blunt the high stresses there. You have no ductility, so you absolutely must get your stresses down. Second, you may not get those impact values. They were generated with fairly thin specimens, much thinner than your half inch thickness. There is a critical thickness where impact resistance suddenly drops because the stress state changes from plane stress to plane strain. This facilitates craze formation, which results is a loss in ductility and impact resistance. Third, your Young's modulus is highly dependent on fiber orientation, which is determined by your flow direction. This is in turn determined by gate location. Orientation is in the flow direction at the surface but can be transverse near the center. And with the half inch thickness, the transverse section could be quite thick. Your material properties will be highly anisotropic. If I was running FEA with a linear-elastic isotropic material model, I would use a value for unfilled ABS for Young's modulus for analysis purposes. Fourth, in the context of plastics, notch sensitivity means that materials that normally fail by ductile rupture will fail by brittle rupture in the presence of a notch. Fifth, if you have any cored out features, such as holes for attaching the trucks, you will get knit lines. The fibers will not span the knit lines, so the strength there will be nowhere near your datasheet values.

1/2" thick, really? Is anybody out there injection molding a solid half inch skateboard? Have you talked to your molder about this? A half inch wall will take forever to cool, so your cycle time, which is mostly cooling, will be huge. Plus, that's what, 4 lbs of material? This will be an expensive part. And good luck controlling shrinkage and warp.

Rick Fischer
Principal Engineer
Argonne National Laboratory

 

[eclipseband]

Thanks Rick,

You've got me thinking a lot...and worried. From your expertise do you think my product will work anywhere near what I am simulating?

I put some wheels and bolts on it with the same material properties as before and was getting FoS of 1.2 in tiny specs around the bolts. Though what you are saying about the fiber orientation has got me worried. Should I be running this with just standard ABS properties? The molding gates will likely run coming from the bottom of the board which screws with the fiber orientation, making it parallel to the force.

There are a lot of boards out there that are even thicker than mine. At my thickest point I am 0.5". My skateboard that I own is around 0.7" all around. This just requires to be done before on large machines. Each board is about 1.6 pounds and quotes I have gotten are a few dollars per part which is fine, I am still projecting great margins, though its not pennies on the dollar like other parts. The tolerances on my part are not a huge issue so unless the warping is too much I am hoping that it is alright.

I appreciate the help.

 

[GregLocock]

Bear in mind that normally the forces on the trucks are compressive and the bolts are actually relaxed by those loads. So your loads and or boundary conditions or contacts must be wrong.

Cheers

Greg Locock


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

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

 

[eclipseband]

Hey Greg,

I think why you're saying it's wrong is probably because I am loading it by placing the load in the middle? Normally riders would have their feet over the trucks which would make compressive forces but I don't think that failure would be an issue for that type of circumstance. I'm trying to test some extreme loading conditions. You never know how people are going to use these things.

 

[rickfischer51]

When I do the math, I get:

W = dens * V = dens * W * L * t = spgr(abs) * dens(h20) * W * L * t = 1.05 * .03611 lbf/in3 * 8" * 32" * .7" = 6.79 lbs.

That assumes density of unfilled ABS. If your board is only 1.6 lbs then it is either highly filled with a low density filler or its foamed. Take the trucks off and throw it in water. I'll bet it floats (maybe even with the trucks on!)



Rick Fischer
Principal Engineer
Argonne National Laboratory

 

[eclipseband]

My material density is 1020 kg/m^3 (0.0368 lbf/in^3). I'll just let you know, it's 70% C PP and 30% Hemp Fiber (Is that considered highly filled?). I was inspired by the Henry Ford hemp plastic car and wanted to bring it back by making a cool consumer product.

My board should be lighter than my competitors and have ecological benefits. I thought that I would have better strength but I am concerned about the effect fiber orientation now. If it's "highly filled" would that result in more random disbursement or even more anisotropic?

 

[rickfischer51]

OK, so W = .0368 lbf/in3 * 8" * 32" * .5" = 4.12 lbs. This is an estimate based on 8 x 32 skateboard blanks commonly available on the internet. How are you getting 1.6 lbs?

Rick Fischer
Principal Engineer
Argonne National Laboratory

 

[eclipseband]

Here is a picture of the type of skateboard I am trying to create:

https://i.ytimg.com/vi/XWBmuqNzBCE/maxresdefault.jpg

This is the "Bantam" by Globe I was referring to earlier where its max thickness is about 0.7". Similarly you can find boards by "Penny" which are even smaller though they have great load taking capability. There are videos of cars rolling over the Penny board without it breaking, but they use their own special compound material.

My board dimensions are 24" x 6.5" x (0.5 max, the geometry has curvature so you could say about) 0.4" x 0.0368 = 2.29lb compared to solidworks estimate of 1.64lb which includes all of the cuts for texture on top and other geometry. I'm quite sure this weight calc is right. The model should be good, it's the FEA that I'm new to and trying to learn.