Dear Doc: I attended your grinding course, where we all chanted The Grinder's Mantra: "Big and dull bad, small and sharp good. Big and dull bad, small and sharp good." But isn't it true that large grits can remove material more quickly than small grits?

The Doc replies: Yes, all things being equal, large grits can remove material more quickly than small grits. But things aren't always equal. The big-dull-bad/small-sharp-good mantra reminds grinders to avoid dressing big grits (say, 60 mesh) dull in order to get a better surface finish – instead of using smaller grits (say, 120 mesh) and dressing them sharp. A dull 60-mesh wheel will give you the same surface finish as a sharp 120-mesh wheel. But a 60-mesh dull wheel will give you high heat generation and exponentially high normal forces. It's difficult to achieve high removal rates with high heat generation and high normal forces.

What about 60-mesh sharp versus 120-mesh sharp? For sure, you'll be able to remove material more quickly with less wheel wear with the 60-mesh-sharp wheel. But 60-mesh dull versus 120-mesh sharp? You'll be able to remove material more quickly with the 120-mesh sharp wheel.

Overhead dresser pros and cons

Dear Doc: We grind nickel alloys in the aerospace industry. Our surface grinder has both an overhead plunge-roll dresser and a table-mounted plunge-roll dresser. I want our machine operators to use the overhead dresser because, well, because we paid for it, and because it's quicker. But they insist on using the table-mounted dresser. Why?

The Doc replies: Let's take the pros and cons of overhead dressers.

First, the cons:

1. Overhead dressers are not near as stiff as table-mounted dressers. That big, heavy diamond roll spindle assembly is mounted on a giant cantilever sticking out over your wheel. It's just not going to be as stiff as a table-mounted dresser. And lack of stiffness means deflection, lower natural frequencies (which are bad) and high chatter risk during dressing.
2. The Grinder's Mantra "Dress and grind at the same spot" applies. In general, you want to dress your wheel at the same point where you grind with it. Doing this makes your wheel egg-shaped in just the right way, which corrects for imbalance (even slight imbalance), weird "2N" oscillations inside the bearings, and other anomalies. Dressing at 180° doesn't correct anomalies. In fact, it can double them.

Now, the pros:

1. With a table-mounted dresser, the wheel has to travel over to the diamond roll, which takes time. If you're dressing once every 20 minutes, this isn't a big issue. If you're dressing once every minute, this is a big issue: You're sucking up cycle time.
2. Overhead dressers can do continuous-dress grinding; table-mounted dressers can't.

Avoiding natural frequencies

Dear Doc: We do large-part cylindrical-traverse grinding and think we may be hitting a spindle natural frequency. Is there a way to figure out the natural frequency without hooking up accelerometers?

The Doc replies: Here's a method that sometimes works, and takes only about 20 minutes. Add some slight imbalance to your wheel. Then, put on some ear plugs, stand on a piece of foam, and place your hand on top of the spindle assembly. Have the operator slowly increase the rpm over the course of several minutes in a smooth motion. For example: 1,500 rpm … 1,550 rpm … 1,600 rpm … 1,650 rpm … all the way to 2,500 rpm. Spindle displacement varies with the imbalance and the square of the rpm – i.e., Displacement equals Imbalance multiplied by rpm2.

Therefore, you're going to feel it increase quickly. If you have a natural frequency in that range, here's what you might feel: More vibration … more vibration … more vibration … A LOT more vibration ... a little less vibration … more vibration … more vibration.

That peak – if you can feel it – is your natural frequency. Let's say it was at 1,670 rpm. Now repeat the test, having the operator go even slower between 1,600 rpm and 1,700 rpm. Try to narrow it down within that range.

If you find the peak again – at, say, 1,654 rpm, you want to avoid running near that rpm. The 1,654 rpm equals a natural frequency of 27.6 Hz.

What if you switch to a heavier wheel? You'll have to repeat the test. Natural frequency decreases as wheel mass increases. But things get even more complicated. When the wheel makes contact with the workpiece, the natural frequency increases.

If this test works – and you discover which spindle rpm to avoid – you will save yourself a lot of frustration.