During the early stages of design, many proposals and inquires were made into central hinge steering. We wanted to try to keep this type of steering because it worked well for us last year, it was easy to design, and it was a unique design compared with the other buggies at the competition. However, this type of steering creates an area along the frame where only one bar connects the two halves of the buggy. This is where last year's buggy failed. This past summer we designed a way to connect the two halves of the buggy using more than one beam. Although we decided against this method of steering, I mention this for future teams to maybe look into this alternative for implementing a central hinge design.
Ultimately, our decision was to try a completely different type of steering, car steering. Car steering consists of wheels that pivot about a point close to the wheel. The wheels pivot independently of the frame and for the most part stay parallel to each other. To visualize what I am talking about think about a car and how its wheels rotate and create an angle between the frame and the wheel. Applying a force at an offset from the wheel axis creates a moment that causes the rotation of the wheels. So my mission was to find a way to apply this force. In almost all cars this force is applied by pushing and pulling on a bar called a tie rod. However the method of applying a force to the tie rods differ in vehicle designs.
While most cars use a rack and pinion system to apply the force to the tie rods, we decided that there were easier ways for us to accomplish this same objective. Also, most of the people in the group were not in favor of having the rider use a steering wheel to turn. The group was split between having the rider use handlebars or using a lever system. If the group decided to utilize handlebars, then we would probably have used a rack and pinion steering design. However, we decided to go with a lever steering method, so I was put in charge of designing a unique type of steering.
The problem for the team was to design a way to convert motion in one direction (the direction the levers move) to motion in a perpendicular direction (the direction the tie rods move). The solution for this problem is to use circular motion. Applying a force tangent to a circle will cause the point 90 degrees away to move in a direction perpendicular to the force applied.
Another major problem for the steering involved allowing for the suspension travel. I wanted to avoid having the situation where a large suspension travel would cause steering. The only way to accomplish this is to have the tie rods lying along a line connecting the center of the circle to the arc created by the suspension travel. The suspension creates an arc around two different points, the top part of the upright (where the top a-arm connects), and the bottom part of the upright (where the bottom a-arm connects)(note: the suspension also creates apparent arcs around all the points between). So if the tie rod travels along one of these radii, movements in suspension should not activate steering.
To create a working design I created a model made out of straws. The model helped me decide how exactly I should implement the levers and where exactly the tie rods should go. The model made it easy for me to visualize and I was able to come up with a good design. After bring the model to a meeting some basic changes were made to it, and the overall design was set. This design consisted of having two levers mounted along the sides of the frame, a rotating piece mounted along the bottom of triangular frame, rods connecting this rotating piece to the levers, two tie rods that connect the wheel upright with the rotating piece.
The critical component for the steering is the piece that I call the rotator. It is a t-shaped bracket, which, as it name implies, pivots about the point where the bars meet in the x-y plane. At the bottom of the 'T', two tie rods attach and go out to the wheels. These two rods will be attached using rod ends so as to allow for misalignment in the z direction in addition to being able to pivot in the x-y plane. The pivoting movement of the rotator causes a pushing or pulling action on the tie rods. For example, if the rotator pivots 30 degrees, then the tie rods should move in the x direction a distance equal to sin(30)*the length of the rotator.
After being moved by the rotator, the tie rods should apply a force on the uprights that connect to the wheels. The tie rods are connected to a toe link that is rigidly attached to the upright at a distance away from the wheel pivot. So that when a force is applied to this bar by the tie rod, there should be a moment created on the wheel. This moment will be what turns the wheel.
A lever system was designed to pivot the rotator. This lever system includes a large handle that pivots against the sides of the frame and a connector bar that connects the bottom of the handle to the rotator. Rod ends will be used at both ends of the connector bar to allow for movement in all directions. The handle bar can be attached to the frame using a pin. In order for the handlebars not to interfere with the rider, the top portion of the bars has to be bent forward and outwards. Click here to download a Matlab .m file that calculates the length of the different bars in the steering system for given input parameters (i.e. desired turn angle, desired hand motion).
One of the major problems facing the Moonbuggy team was designing our systems to work together. I had a great deal of trouble trying to fit the steering design on the buggy and not interfering with the rider or the suspension. After the team created a full scale wooden model of the buggy, I decided that I would mount the rotator below the frame. This would allow me to have my connecting rods and tie rods all below the frame and suspension arms.
In summary, the steering system works as follows: The rider pushes and pulls on the two handle levers which each pivot about the top frame member. This causes a linear force on the connecting rods which connect to the rotator at its two ends (sides of the 'T') and cause it to pivot. The tie rods are connected to the rotator at its third (top of 'T') end and are translated linearly side to side by the rotation of the rotator. This translation pushes on one upright and pulls on the other causing both wheels to turn in the same direction. Click here to download a Matlab .m file that shows an animation of the steering in action.
Some analysis was done on the members of the steering assembly in order to obtain a value for their thickness. Most of the bars will only be experiencing small amounts of compressive and tensile forces (no bending) and therefore almost any reasonable size tube (1/2" dia.) will work fine. The handle lever however will have a large amount of bending stress because of its length, and its dimensions will have to be a lot thicker.
All connecting rods of the steering assembly were made from thin walled 1/2" x .035" 4130 chromoly tubing. Steel inserts were used to allow for rod ends to be threaded in at the end of each rod. Rod ends allow a connection to have two degrees of rotational freedom. All our rod ends were ¼-20 thread and were obtained from National Rod Ends (http:\\www.Nationalrodends.com). The rod ends were connected to the rods via a threaded insert that was welded onto each rod. Each connecting rod was made with one right and one left-handed thread insert. This design enabled the rod to be lengthened and shortened by turning the rod. However because of time constraints and lack of supplies, the connections between the handlebars and the rotator was made with two right hand rod ends. This linkage is the least critical as far as requiring adjustability and therefore it did not create any problems.
The rotator was constructed from two 3/4" square tubes (.035" wall thickness) welded together to form a T-shape. The rotator was fabricated from square tubing because it was simple to make and lightweight. Square tubing also allowed for the rod ends to sit inside slots cut into the sides of the tubing. This meant that the bolt connecting the rod ends to the rotator would be in double shear. Double shear helps to minimize the stress in the bolt and the bearing stress on the rotator. A long bolt connected the rotator to the bottom tube of the frame. A spacer was placed in between the rotator and the frame to lower the rotator. This was done to make sure that the steering assembly would not interfere with the A-arms. A thrust bearing was also placed in between the spacer and the rotator. This allowed us to keep the bolt connecting the rotator to the frame tight while still providing little friction to the piece as it rotates.
A U-channel shaped piece was also welded onto the uprights to provide the leverage required to turn the wheels. This piece, called a toe link, was welded at an angle known as Ackerman's angle. Ackerman's angle attempts to utilize steering geometry in order to turn the wheels the proper amount during a turn. During turning the inner and outer wheels need to be at different angles because each wheel is a different distance from the turn center. Calculations can be made to find the correct angle necessary to achieve Ackerman steering based upon the wheel base and front axle tread. However, because of time constraints, these calculations were never made and the angle was estimated based upon a simple formula which states the toe link should point at the center of the rear axle. The complete formula needed to solve for Ackerman angle can be found in the book: 'The Truck Steering System form Hand Wheel to Road Wheel' by J. W. Durstine. Even though we used a simple estimation for the angle, the Ackerman steering design appears to have worked as required during the competition.
Because of the strength needed and the shape desired, we decided to make the handlebars out of carbon fiber. To do this we split the handle bar up into two halves-half made from 1" square steel tubing and the other made from carbon fiber. The bottom half consisted of three pieces of steel tubing welded together in a shape that bends towards the center of the buggy where it connects to a rod end in a double shear set up. An aluminum piece was fabricated to fit snugly inside the steel tubing at its top end, and it was used to connect the two halves. A bolt was placed through the steel tubing, aluminum piece, and the frame to hold the two halves together and connect the handlebars to the buggy. A long light weight aluminum tube of 1/4" outer diameter was then welded to the top end of the aluminum piece. The tube was very thin and could be easily bent into the required shape for the handlebars.
Prof. Petrina helped advise us with the construction of the carbon fiber. After doing some calculations and constructing a test specimen, it was decided that fifteen longitudinally placed layers of carbon fiber were required. Using prefabricated carbon fiber strips that the professor donated to our team, we wrapped the carbon fiber around the thin aluminum tube. One handlebar was wrapped with the carbon fiber before the tube was bent to its final shape while the other handle bar was bent prior to placing the carbon fiber on it. Placing the carbon fiber onto the already bent tube proved to be very difficult and time consuming. On the otherhand, the placement of carbon fiber onto a straight tube was very easy, but when the tube was bent the carbon fiber began to kink. However fixing the kinks in the carbon fiber was not difficult and therefore I would recommend that the future tubes should be bent after the carbon fiber was applied.
The day before the competition, the carbon fiber came off of the aluminum piece that it was connected to. During fabrication, we were concerned with how well the carbon fiber would bond to the aluminum but Prof. Petrina told us that it should not be a problem. But as we were loading the buggy into the van, the carbon fiber slid right off. Fortunately the carbon fiber did not break but only slid off of the aluminum. To fix this problem we placed epoxy on the surface connecting the carbon fiber to the aluminum. We then placed a bolt right through the carbon fiber and aluminum for some added support. After these changes were made, the carbon fiber handlebars acted exactly as designed during the competition. Since that time, the epoxy that was smeared around the outside of the carbon and aluminum has shown signs of flaking, however that is no reason to believe that the epoxy between the two surfaces is loosening.
At the competition the overall performance of the steering system was acceptable. The turning radius of the buggy was well within what was at the race, and the driver never lost control of the buggy. The driver did however encounter some problems while going over obstacles. The large uneven bumps caused the steering to activate, and the driver had problems preventing this. Eventually the driver learned that he could wedge his elbows against the side of his body as he held the handles, which enabled him to keep the handles from moving.
After riding the buggy for the first time, it was quite obvious that the steering system was to sensitive to the position of the handlebars. When it was initially designed, I over estimated the amount of degrees the wheel needed to move in order to make a turn. Because of this incorrect estimate, the rider would only have to move his hands forward one inch for the buggy to go into a undesirably sharp turn. To fix this problem the geometry of the steering system had to be changed. So between our two runs at the competition, we effectively shorted the length of the rotator by welding two plates onto it and moving the rod end holes closer to the pivot. This adjustment gave the driver more control, and it seemed to work better during the second run. Although the steering is now acceptable, it still is more sensitive than desired. So this problem would definitely have to be fixed if the buggy is going to be used in other competitions.
Although the system worked decently, it still needs some changes if it is going to be used in future years. Making the system less sensitive (i.e. increasing the driver's mechanical advantage) and redesigning the way the rotator connects to the frame are two suggestions for ways to improve the steering system. I also recommend that future team members research other types of steering systems such as rack and pinion steering or using hydraulic cylinders. Our steering worked well but other systems might render even better.