Learning Forum

[Victor Silk]

Hi all! I'm a novice in ANSYS, so I'm confused on how to approach the following problem. I need to analyse the effect of a moving load-carrying box on a gravity roller conveyor. Can someone please help me with the steps?
P.S. The model contains only 10 rollers and a box

[peteroznewman]

What information do you want the model to provide? Depending on what you want from the model, there are different models you could build.

[Victor Silk]

I need the values of the stress and deformation induced on the rollers due to the weight of the box. I don't know what kind of contacts, connections and settings to give. It's just a college project.

[peteroznewman]

Hi Victor,

It's a good college project.  Are you given the geometry or do you need to create it? If you need to create some geometry, do you have a CAD system that you're skilled at?  If so, you can create the geometry in that and export a neutral file for import into SpaceClaim.

If you are not good at a CAD system, you can create the geometry in SpaceClaim.  You will have to look at some of the tutorials on this site.

For the rollers, I expect they are constructed from tubes with a given outer diameter and wall thickness. What are those values? What material is the tube made of?  You will need the Young's modulus and Poisson's Ratio of that material. The rollers have some bearing at each end. What is the distance between the bearings? What is the spacing between the roller centerlines?  What is the angle the plane of rollers are tilted at relative to horizontal?  What is the tolerance on the roller diameter?  What is the tolerance of the roller centerline deviation out of the plane?

For the box, what are the dimensions of the base of the box?  What is the weight of the box? Are you going to simulate the flexibility of the box? You could simulate a rigid box, or a flexible box. Is the bottom of the box flat or is it convex?

Engineers often design for the worst case condition. In this case, if you had a box with a convex bottom that was fairly stiff, and you had one roller out of the ten that was a little larger and mounted higher out of the plane of the rollers, you could do a worst case stress and deformation by loading the center of one roller with the entire weight of the box. That way you don't need to construct anything but a single roller. You can even do half a roller and use a symmetry boundary condition so that you can apply half the load at the centerplane where the geometry was cut.

Draw a circle at the average radius of the outer and inner walls on the YZ plane, that being the center of the length of the roller. Extrude that circle along the X axis for half the length between the bearings. That cylindrical surface can be your tube to which you can assign a wall thickness that goes equally on both sides of the circle. The edge of the cylinder in the YZ plane is assigned an X=0 displacement constraint.  The other edge of the cylinder where the bearing goes has a remote displacement with X free, Y=0 and Z=0 displacements applied and rotation about X set to zero.  You will want to split the circle in the XY plane in order to get a vertex at the top of the roller where the load will be applied in the downward direction (-Y for example) corresponding to half the weight of the box.

 

[Victor Silk]

Thanks a lot Peteroznewman. You seem to be an expert in this. But the problem is not at engineer's level. I have designed the solidworks model of the rollers with ball bearings on either side for all the rollers and the is positioned tangentially to the rollers. Since I'm a beginner in ANSYS, I would like to do a basic deformation and stress analysis. The fact is I don't know whether to go with static structural or transient structural and also how to specify the contacts, fixtures, joints, etc. I just couldn't find a proper tutorial video for this sort of problem.

[peteroznewman]

Hi Victor,

There is a tremendous amount of detail required to design parts to make a working physical system.

Contrast that with a simple analysis that predicts that the critical component of that system will not fail.

The weight of the box is a design requirement that your design must meet. The roller bearing is a component that you will select that has the capacity to support the load. You are not designing the roller bearing. You will select the right size out of a catalog.

You are designing the tube that the roller bearing will be mounted in. You must show that the tube will not fail when the weight of the box is applied. You don't need all the detail of the roller bearing present in a model to determine if the tube will fail. If this were not a college project, you might also be designing the frame the holds the rollers, so that would be another analysis that you would do.

The art of analysis is deciding how to idealize a problem to take out the irrelevant details and leave in the critical component where there response is unknown. In your case, that would be the tube, which is why I described a very simple model that answers the question of whether the tube will support the load or not.

A basic deformation and stress in a static structural analysis of the tube is a good place to begin to learn how to use ANSYS. 

[Victor Silk]

Thanks again, Sir.
So, I cannot evaluate the results while the box is moving on the rollers, can I?

[peteroznewman]

Say you wanted to design the rail at the end of the slide that stops the box. The box of mass M travelling at velocity V impacts the center of that rail, then you might build a Transient Dynamics model to evaluate if the rail will collapse, and design one that doesn't.

If you are designing the rollers so that the simple worst case of all the weight of the box is on one roller, then you don't have to evaluate a more complicated case. Yes, you could build a model where the box is moving on the rollers, but why would you?

But if you say the box of mass M is dropped from a height H onto the roller, then a more complicated Transient Dynamics model might be warranted. You see it all depends on the definition and scope of the problem.

 

[Victor Silk]

Thanks. So static analysis is enough to deal with roller design, is it? How about the bearings?

[peteroznewman]

Engineered products work by being able to analyze a system and break it down into components that will be purchased and those that will be custom designed. An analysis calculates the loads that components must support. Once you have specified the load on a component you use that information to select the correct size from the catalog.

If you are designing a roller conveyor system, the bearings are a purchased component.  The bearing manufacturers make it very easy for you to select a bearing that will have the required life under the specified conditions of load and speed. You design the parts that mount and retain the bearing according to the manufacturer's specifications, then you are done. No analysis required. The bearing manufacturer has done that for you. You spend your time analyzing the custom designed parts, and to be efficient with your time, you use the simplest model that will give you an acceptable design. A static analysis is sufficient to design the roller.

If you want to be a bearing design engineer, then of course you will be doing a lot of analysis about roller bearings, but you won't care about a roller conveyor system. 

[Victor Silk]

Thanks a lot peterzonewman. You seem to be the only one who is helping out all beginners like me. Thanks for all the facts which none of my college faculties came out with. You did give me a good start. For now I have done the static analysis on the rollers. I will have to repeat the process to minimize the cost involved.
Thank you so much ?