Someone told me to avoid over-sizing when using DC Motors. Why? Is this because of sparking and higher wear (brushes) while the motor is running at a lower power rating?
Can someone explain?
This answer assumes that you meant to ask about over-sizing a DC motor rather than over-dimensioning it. over-dimensioning is when you provide too many dimensions such that they might be in conflict with each other based on drafting errors, rounding, or tolerances. For example if you draw a 3 story building and indicate that each floor is to be 10 feet tall, but you also dimension that the overall building must be 31 feet tall, nobody would know where to put the extra foot.
Over-sizing is when you pick an item that is bigger than you theoretically need. For example if structural analysis indicates that you need a 12" tall beam, but you decide to specify a 14" beam to be safe.
Over-sizing motors is common - it is typical in the mechanical world to specify a motor 25 - 50% larger than your calculations show you need as a rule of thumb. Over-sizing motors is useful because there could be some variation in the load the motor supports. A pump may have to run a little faster than it was designed to, or a bearing could be improperly lubricated, increasing friction, for example. In this case, it is desirable for the motor to be able to function even though conditions are not exactly as designed. It is also usually costly to replace a motor with a larger motor, so the relatively small cost of buying a larger motor in the first place is usually worth it.
There are also some down sides to over-sizing motors though, which is why it's unusual to see a motor 100% or 200% bigger than it theoretically needs to be. One obvious issue is cost. Bigger motors are more expensive and buying a bigger motor than you ever need means spending money that didn't have to be spent. There's also an issue with efficiency. Most decently sized motors are more efficient in the top half of their speed range. If they are used at very low speeds, they will waste more energy as heat. This is compounded by the fact that there is more rotating mass in the motor, so it takes more energy to start and stop the machine. In addition, some larger fans have cooling fans attached to the motor shaft to blow cool air across the motor, preventing overheating. At very low speeds, this fan does not move quickly enough to provide cooling air, so the motor is more likely to overheat which will damage the motor.
One other wrinkle is that variable speed controllers have limited resolution, and if your useful range for the motor is all in the lowest 10% of its speed range, your controller might not be able to regulate speed as carefully as you need it to. This is a constraint that can be designed around, but adds complexity and limitations to your system.
So motor size is always a trade-off between the risks of having a motor that is too small to move the load and a motor that is too big and wastes power or overheats. Most industries have settled in the 25 - 50% range but there are certainly valid exceptions where efficiency or reliability are strongly controlling factors.
This Farnell datasheet has a good overview of these principles, and most motor vendors also have tools or guides that will help you with motor sizing. Some even publish efficiency curves for their motors.
Overdimensioning is never good because there is a problem called, "tolerance stack up". Tolerance stack up is essentially predicament where you consider way too much tolerances and thus throwing everything out of proportion. In other words, too much modification can create a large amount of ambiguity.
For instance, if I wanted to create a particular amount of magnetism in the motor, I have change "x". Okay, now that I have changed "x", I now need to change "y" because "x" changed. Well now I have to change "z" because of "y", et cetera, et cetera.
This is why we learn about optimization in Calculus and Matrix Algebra in order to combat that problem :)