BRAK Project-> Design info-> Mechanics-> Base goals
Mobile Base Design Goals
When we first started designing the base, the goals were simple: a robot on which we could do mobile robotics research. You wouldn't believe how many robots can be designed which meet those specifications. So, as we began to deal with reality, these were some of the design decisions and objectives for the BRAK Project:- Use one motor for steering, one for transmission As mentioned earlier, one of the complexities of the synchro design is assuring the steering and turning components move at the same rate (and the same distance, too). It was decided early in the second stage of the design to use only two motors, one to steer and the other to drive. (The first design in 1992 used one steering motor but three drive motors, each located within the wheel housing. I abandoned this idea when it became apparent that stepper motors of sufficient torque would be too heavy and consume too much power while geared DC motors would be difficult to keep synchronized). This makes it easier to guarantee synchronization but does so by moving the complexity issue to designing mechanisms to transfer the power from the motor to the wheels.
- Use assembled gearboxes DC motors typically have their maximum torque at high RPMs (often anywhere from 2,000 to 15,000 RPM); this maximum torque typically is also measured in ounces per inch at the size and DC voltage (less than 24VDC) suitable for a mobile robot. The robot needs lots of torque (in pounds per inch) but at lower RPMs (for example, 30 to 120 RPM). Result: the motor output need to be geared down to the slower speeds. Gearbox design is an art to itself, and designing and building your own can be time-consuming and expensive. I opted to take the advice of others and (1) look for small gearboxes which could be easily designed into the robot or (2) look for DC motors with gear reduction built in.
- Use chain-type sprockets/belts to reduce slippage With the appropriate motor/gearbox combination the power still has to be transferred to the three drive and/or transmission shafts. There are basically two choices at this point: (1) gears or (2) sprockets and belts (and yes, I know, I just tried to scare you away from using gears... well, this is the real world). Both have their advantages but the major disadvantages from my standpoint is gears are usually expensive and have this characteristic called backlash while sprockets and chains can stretch or slip (which is the evil equivalent of backlash). However, there are chain-like sprockets and chains which virtually eliminate slippage. They are more expensive but if accuracy is of importance they're the way to go.
- Symmetrical design for machining and weight distribution I tried to make the design as symmetric as possible. Synchro designs are typically three-wheel designs so equilateral triangles tend to be prominent in the base's geometry, but for other little things -- like the placement of access holes or mounting holes -- I think it makes the actual construction easier since is requires less set up in the machine shop. Symmetry also is useful in trying to maintain the weight distribution of the base; placing the motors symmetrically balances the design (it also good to think about things like where the batteries would be placed, where cables and cords will be routed, where you need access to screws or bolts, etc).
- Make the main framework as strong as possible, build from there In addition to my concern for an accurate and powerful drive and steering system, I was worried about the strength of the base itself. My original goal was for a platform/payload weighing 50 pounds. I admit my ignorance in the strength of materials (Brak is really over-built, I'm told) but I wanted to be sure the main structure of the base would withstand the weight of the platform/payload without flexing and be resistant to torsional forces as well. With a strong framework, lighter materials can be used from there for the rest of Thor structure.
- Use cheap parts where not critical There are places you can cut corners and places you can't; I decided early that the accuracy and strength of the base were the most important things in the design and I wouldn't skimp on them. In less crucial areas it was OK to scrimp -- and it's not a bad thing to scrimp when you can find a cheaper way to do the same jobs that are crucial.
- Use duplicate parts Duplicate parts in the design are a good thing. Always buy more than you need anyway, but it you re-use the same part in the design it has other advantages: (a) if you loose it or break it, you've got replacements, (b) many parts are only sold in packs of 10, 24, or 100, (c) many parts which you can buy onesey-twosey are cheaper per part in quantities of 100, and (d) the industry buzzword for it is standardization.
- Make easy to disassemble and easy to access Again, I'm not a mechanical engineer (of course, if you're in robotics you really should be, and a electrical engineer, and computer engineer or computer scientist, or at any rate you're likely to end up know how to build the same things but without understanding any of the math behind the dynamics, but I digress) but it stands to reason that any time you don't think about how you're going to get a screwdriver into some area that will be the most crucial area to access with said torque tool
- Not worry about minimizing weight yet Literally the last concern I had was for minimizing weight. This is not to say if the design was going to weigh 100 pounds that was OK; it means that once I was sure the design was strong (and worked) I'd go back and see where it might be lightened (it will start with mbase2, I'm sure).
