Realistic loadings for design

Do you have access to FEA software? If you do, I'd suggest maybe picking a target weight, and optimising the structure at a given weight to be as strong as possible as evenly as possible.

dilzy said:

Yeah mate, it's actually for an FEA assignment. I'll be using Ansys probably classic rather than workbench because I'm much more familiar with it.

Not a bad idea, but it'd be nice to know what people in the industry actually use. I swear specialized don't do side loadings on their swingarms.

Click to expand...

I'd be surprised if more than 10% of the bike industry did any rated-load testing or simulation at all. Specialized reckon they do, as do Trek, yet neither of them have substantially better structural reliability than anyone else does in my experience. Of course, I can't prove what I suspect.

Exactly what are the criteria for the assignment you have?

Maybe you could try to contact the German BikeMagazin.de. They do execute longitudinal & torsion stiffness tests when reviewing front-susp. forks and rear-susp. linkages. I'd say that another company does that for them. Maybe they could give you some contacts and/or descriptions. Some descriptions of their work here.

I think a lot of companies just use Solidworks and then use the Cosmos Express FEA which is built into SW.

I seriously doubt there are any "benchmark" like numbers out there as things get much more dicey when talking about complex dynamic loads. And since we are talking about the bike industry, aside from some of the larger companies most are probably not going through a proper iterative process to determine load standards or proper design safety factors.

I don't know how large a project this is supposed to be, but you should be able to determine load scenarios through simple calculations and that should give you a decent gauge of where abouts your values should be. A more refined approach would be the application of data acquisition components (strain gauges or load cells, etc), as well as destructive analysis with tools like an Instron tester. But all the variables such as geometries/speeds/deceleration/acceleration/delta t/masses are all right there for you to use.

start with a free body diagram of the bike and your body weight. Assume a certain G load or calulate the G load to bottom your suspension. Add a 2X safety factor and call it good. I would bet this project is to teach you to run the FEA software rather than a benchmarking/ reverse engineering exercise.

I'm not so sure how willing they would be to share their information, but Cannondale in particular does some really intense testing on their bikes. I remember seeing something about their test bikes, loaded up with strain gauges feeding to a backpack mounted collection system.

You could always do a few quick calculations, take your weight dropping from some height, find out the force it exerts. FEA in your case will be more about optimizing the design rather than trying to perfectly simulate a real world situation. I learned that quickly when I had to do hundreds of iterations on my design project in school.

ChrisKring said:

start with a free body diagram of the bike and your body weight. Assume a certain G load or calulate the G load to bottom your suspension. Add a 2X safety factor and call it good. I would bet this project is to teach you to run the FEA software rather than a benchmarking/ reverse engineering exercise.

Click to expand...

That assumes you only ever just barely bottom your suspension out - hard bottom outs, or even super fast hits where the damping force is considerably higher than the spring force, would definitely exceed 2x the bottom out force IMO. I'd go for a factor of safety more like 10.

You are probably right that the actual force is higher than 2x bottom out force. The safety factor should probably be higher.

My basic calcs for my bike:

500 lb/in spring X 2.75 in stroke X 1"/3" = 458 pounds Obviously, I negated damping.

What is the average amount that someone can squat? 2-3 X body weight? So for a 200 pound rider, you would be into 600 - 800 pounds of force. The force for a breif moment would probably be higher.

Eitherway, i don't think that an exact number is the goal of the exercise.

I've been thinking about this problem recently as well. What happens if you were to theoretically put a spring twice as stiff on your bike (1000lb/in rather than 500lb/in). Perhaps turn the dampening all the way down. Would you be able to bottom it? I would guess not but I don't know, I've never tried it. Is this a valid analogy? I'm not sure it is, but something to think about perhaps.

I agree that 2x is probably a good starting point; I would be surprised if it was as high as 10x but maybe... Someone needs to strain gauge their bike and get back to us.

if you are assuming a failure due to lateral load at the rear wheel, you could start with the lateral load that a wheel fails at.

Most real world frame failures I see are:

1) the head tube to down tube and top tube joint failing in a bottom out type load. It is highly likely from multiple impacts

or

2) failures in the swingarm caused by pre-existing structural damage.

or

3) failures in the seat tube around swingarm / shock mounting joints.

Another approach you can use is benchmarking. You could model existing bikes and run a FEA ito determine the failure loads. The apply these loads to your design.

RMboy said:

I think a lot of companies just use Solidworks and then use the Cosmos Express FEA which is built into SW.

Click to expand...

I like SolidWorks, but I hope companies are not using Cosmos Express as it is crap. The full version of Cosmos is decent, but not as good as Ansys.

Internet engineering is damn funny. Run some tests buddy.

dilzy said:

Bottom-out loads are not the ones that kill bikes,

Click to expand...

Yes, they do.

RMboy said:

I think a lot of companies just use Solidworks and then use the Cosmos Express FEA which is built into SW.

Click to expand...

I hope they splurge for at least the whole CosmosWorks Package... but when you hafta pay for it (and not just use it on one of the university's comp's) it get's expensive I'd imagine.

And for some bottom out's I'd say using how much you can squat wouldn't be adequate, like the ones where your just glad your seat kept you from getting violated by the seat post you just compressed all the way even though you thought it was clamped tight lol.

See if you can track down some standards required by the CPSC. We have to do a test for our seatpost that is exhaustively detailed. The standards are really only meant for regular bikes, but it'll give you a good starting point.