Steve Blackmore wrote: >> Have you considered a toothed-belt actuator for either the knee or the >> spindle? I saw a very neat one which used a second belt bonded to the >> ways to eliminate belt-stretch problems. > > What belt stretch? Once they are adjusted up, that's it.
When using a belt to transfer rotary motion from one pulley to another, belt stretch is usually negligible. That is very much NOT the case when using a toothed belt to convert rotary motion to linear motion. > Timing belts have steel strands in them, to add strength and ensure they > don't stretch. Everything stretches, including steel. Timing belts have very thin steel strands. Assume a type L timing belt (3/8" tooth pitch) that is 1" wide. The overall height of an L belt is 0.140 inches. The teeth are 0.075 high, leaving only 0.065 for the tension members. Assume that one-half of that 1" x 0.065" area is steel - the rest is rubber, etc. (I bet the actual percentage of steel is a lot lower - maybe 25%. There is space between the individual wires that make up each strand, and space above, below, and between the strands.) With the optimistic 50% steel estimate, you have 1 * 0.065 * 0.5 = 0.0325 square inches of steel. Assume a load of 200 lbs. That results in stress in the steel of 200/0.0325 = 6153 psi. The tensile modulus of steel is 30,000,000 psi. So under 200 lbs load, the belt will stretch by 6153/30000000 = 0.0002" per inch. If you are using that belt for a 100" long linear axis, the stretch would be 0.020. A ballscrew for an axis that long would most likely be over an inch in diameter. A 1" ballscrew probably has a root diameter of at least 0.8", resulting in 0.502 square inches of steel. That is 15 times more steel, or 15 times less stretch. So a 100" long, 1" diameter ballscrew loaded to 200 lbs would only stretch 0.0013" instead of 0.020". > Example - On most modern OHC motor engines once the > timing belt is fitted and adjusted, it's never touched again for 30-50K > miles. Machine tools and car engines aren't quite the same thing. Assume the engine has a 3" diameter pulley. If the belt stretches by 0.010", that amounts to 0.010/1.5 = 0.00667 radians, or about 0.38 degrees. The valve timing in a car engine doesn't mind 3/8 of a degree of timing error. But 0.010" of linear error is huge for a machine tool. That isn't to say that timing belts can't be used for linear motion, but you have to do the math, and decide how much stretch is too much. Regards, John Kasunich ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
