Not for the weak minded.

Down to Business

June 18th, 2009 Josh
This entry is part 3 of 3 in the series MMA of Engineering

So what exactly do I do? I design mechanisms to work according to the specifications given by my customer.  As an example, I’ll do a simplified design of a kick stand for a motorcycle.

The requirements are that it support a 500 lb motorcycle with the center of mass 2′ from the ground when the bike is standing vertical.  The kick stand must support the bike tilted at an 80 degree angle and have a safety factor of 2 for the ultimate strength of the material. It must be made of Aluminum (specifically 7075-T6 alloy), and be optimized for weight.

Given those requirements the first thing to determine is the load on the kick stand. The 500 lb bike leans at an 80° angle, so the kick stand must support the moment (force times distance) about the point where the tires contact the ground. To find the moment, we need to find the perpendicular distance from the tire contact patch to the line of force that gravity is exerting on the center of mass. That distance is 2′ * cos(80°). To get the moment, we multiply the distance by the force, which is 500 lbf: 2′ * cos(80°) * 500 lbf = 174 ft-lbs. This means that the kick stand must be able to supply 174 ft-lbs of torque to the chassis to keep the bike upright.

To keep things simple, let’s just say that the kick stand contacts the ground 12″ away from the tire. That makes the math easy to find the force on the foot of the kick stand. 174 ft-lbs/1 ft = 174 lbs. I will also have an interface to fasten the kick stand to, and with this, I can start with a general shape for the kick stand. I’ll draw this in my 3D modeling program:

start

blog-job2

blog-job3

blog-job4

Ok, now I have a general kick stand shape that fits the interface on the bike and contacts the ground. From this point, there are two routes to go to analyze the strength. First, I can analyze it with classic hand methods. Second, I can export it to a software package that analyzes it for me. Usually I must do both to verify that the software is giving me reasonable results. Since I am doing this as an example, I’m going to skip the hand calcs which take quite a bit of time and jump straight to the more interesting computer analysis.

When you were in science class as a kid, did you ever use play-dough and sticks or tinker-toys to make molecules? The reason I bring that up is that the analysis software takes my model and turns it into a shape made of tiny nodes and beams (nodes are like the play-dough, and beams are the sticks). blog-job7You can see in the picture how the software recreates the model with these “tinker-toys”. The reason it does this is that for every beam or “stick” the software runs a complex set of equations to determine how that stick reacts to the forces applied to it. If I want to run a quick analysis, I need less sticks, but that also means that it doesn’t look as much like my model as it should. If I have a critical area, I tell it to put more sticks there so I can see exactly what is happening. This makes the time required to solve all the equations much longer. This is where the trade off exists between solving something fast, and getting accurate results.

This model is pretty simple, so I told the program to just size the sticks around .1″ long. That is good enough. I then told the software to hold the kickstand at the bolt hole and press on the bottom with 174 lbs of force.

blog-job5

The program did not take long to run. I had the program display the stress in the material because that cannot exceed the ultimate strength of the material with the factor of safety. Since the factor of safety is 2, the stress cannot go above half of the ultimate strength. For 7075-T6, this is about 76,000 psi. So the maximum stress must be less than 38,000 psi.

blog-job6

According to the software, the maximum stress is 23,300 psi, well under the requirement. In the picture you can see exactly where the stress is. Notice the blue low stress area in the center? This tells me that I can remove all that material with little or no effect on the part. This is how a part is optimized for weight. Also, if you look carefully, you can see that the model is actually bent compared to the first picture. This is a visual aid (magnified to make it more obvious) of how it is bending. There are many other things this program will do… but we’ll stick to the simple stress analysis.

The next step is to go back to the model and change it to remove as much material and adjust the shapes to minimize the sharp corners where stress likes to build up. At the moment it weighs .336 lb.

Here is the revised kick stand:

blog-job9

As you can see, I have cut away most of the blue area, as well as thinning down the foot. This new kick stand weighs .237lb. That’s close to a 30% reduction in weight. Now lets see how it does in the analysis.

blog-job8

Much better! The stresses are higher, but still well under the 38ksi limit. You can see how the areas of low stress have been significantly reduced indicating a successful weight reduction. The red area has spread out due to the radius that I put there. The reason that it has increased is because I took away material behind it in the pocketed region.

For a motorcycle, this is probably good enough. With aircraft, however, weight is the most important factor since each pound saved saves significant fuel costs over the life of the aircraft. In this case, this process would be repeated several times, until the maximum stress within the part is close to 38,000 psi, and as much of the low stress areas have been removed as possible.

This part would then be prototyped and tested in real life before going on to production. There are many other aspects that I must consider in my job, and seldom are parts as easy to analyze as this. There are also many other aspects that I must consider daily. To name a few: corrosion, friction, thermal expansion, errors in manufacturing, as well as considering if someone could accidentally break it, even if it is designed to do it’s job.

In my current project I have designed about 50 unique parts, ranging from simple shafts to complex bodies that have to be designed for multiple loads, vibration, and harsh temperature changes. The stress analysis is just a small slice of my job. I also have to make sure that all of my mechanisms properly move, interface, and are capable of being built.

So that’s a little look into part of what I do. Hopefully you learned a little in the process. Just think, every man-made object you touch has gone through this same process to some extent. Try to count how many different things you touch in a day! That’s why I consider the Mechanical Engineer the ultimate Mixed Martial Arts fighter of the engineering world.