How to run a grinding test (part 1 of 4)

Often, grinding processes do not reach their full potential due to the lack of a systematic approach. The following describes one such systematic approach that may help to establish useful parameters of grinding processes. The recommended approach divides into six process steps. A few simple formulas define some of the critical parameters.

Process steps:

1.      Objectives

2.      Preparation

3.      Running

4.      Evaluation

5.      Recording

6.      Publishing

Objective

First, be clear what you or others want to achieve. Some objectives may conflict with each other, such as aiming for higher removal rates while, at the same time, trying to achieve better process stability. Some of the objectives may be:

1.      Fixing an existing problem (burning, surface finish, dimensional stability)

2.      Better process economy such as a longer wheel life

3.      High material removal rates, i.e., shorter cycle times

4.      A new specification to be tested

5.      More reproducible results

6.      Increased process stability

Preparation

If the task is an existing process, make yourself familiar with the critical issues and record them. These include:

1.      Material removal rate Q’w or Q-prime

2.      Wheel surface speed vc

3.      Speed ratio qs

4.      Feed-rates vf, vfa

5.      The total amount of grinding allowance and depth of cut ae and the number of cuts

6.      Check coolant nozzles for damage, establish and record nozzle position

7.      Measure coolant temperature (morning and evening) and check flow rates

8.      Dressing parameters such as speed ratio qs, dressing infeed ad, dressing cross-feed-rate vd, synchronous or asynchronous dress

9.      Check Amp meter during the grinding cycle, or use a grinding monitor. Record highest reading during roughing and finishing

10.  Check air-grinding time. Remember: Actual grinding time is often only a fraction of the cycle time. Gains in cycle time reduction can often be achieved by reducing air-grinding time rather than by increasing feed-rates. (Air grinding time refers to the time when the wheel is not engaged in cutting but moving around the working envelop of the grinder)

Before running any cycles, also ask the operators about the main problems that they encounter. These questions include:

1.      Does or did the wheel lose form?

2.      Did you experience burning or chatter?

3.      Is the process stable?

4.      How does the coolant supply behave? Is it the same temperature across a full shift?

5.      Is the fixturing rigid? Do vibrations occur?

6.      Can you maintain dimensional stability?

7.      Is the surface finish ok, and is it stable?

Running

If using a vitrified bonded wheel, before mounting the wheel, ring it! Use plastic blotters between the metal flange and the vitrified grinding wheel. Tighten the flange screws with a torque wrench. Unless wheel suppliers recommend any other torque values, use 20 foot-pounds or about 30 Nm to tighten the flange bolts. Superabrasive wheels with metal bodies do not require plastic flanges.

If you feel the current parameters for a given wheel are reasonable, repeat those parameters with a new wheel specification and compare results.

Let us assume you are running a vitrified wheel test, and you have just a few hours available for the test. If so, consider turning down the diameter of the test wheel halfway between the new wheel and the stub wheel diameter. Often, a new wheel gives excellent results. However, as the diameter gets smaller, the grinding performance of the wheel diminishes proportionally. Remember, 100 mm in diameter equals 314 mm in wheel circumference, and hence fewer abrasive grits have to share the same workload.

Evaluation

Before making too many changes to any given process, run it for a while, ideally through the full life of one grinding wheel if you are working with conventional abrasives.

Also, keep in mind that sound engineering is to change only one parameter at a given time, to observe changes, to evaluate, and then to apply changes.

Measure components for dimensional accuracy, burr formation, and grinding abuse (burning). Given that you have knocked out any unnecessary reciprocating passes, increase the feed-rate vfa to shorten cycle time further. To reduce wheel consumption, reduce the dress infeeds ad. Ideally, push the wheel to break down or to burning. For this purpose, particularly when testing for burning, you may require some reject components.

What have you observed? How did the wheel behave? Loss of form indicates one or several of the following things:

1. §The wheel is too soft
2.  Surface speed is too low (increase it by steps of 3 to 5 m/s)
3. Feed-rates are too high
4. Amount of dressing is insufficient
5. The grit size is too coarse

Did the process result in burning?

Check whether the wheel is sharp after dressing. To do this, stop the wheel and feel it with your fingertips. If the wheel feels smooth, your dressing parameters need adjusting, or the wheel needs to be opened with a dressing stick. Check coolant delivery. Is the nozzle in the right position? Does the fixture cut off the coolant flow? Is the surface speed too high?

For this reason, try to reduce the wheel speed by 3 to 5 m/s, as this makes the wheel act softer. You may increase the feed-rate or infeed per revolution to increase the self-sharpening of the wheel. For this purpose, use 10% steps and observe results.

Recording

Create a spreadsheet that lists the main parameters for each of the passes that are needed to grind a component. Record the following parameters:

1.      Wheel speed vc

2.      Depth of cut ae per revolution of the workpiece (OD plunge grinding)

3.      Feed-rate vfa for OD reciprocal grinding

4.      Material rate Q’w

5.      Average chip thickness hm

6.      Dressing feed rate vd and dressing depth ad

7.       Dressing ratio qd and the total amount of dress ad

8.      Number of workpieces between two dressing cycle

9.      Amp load on grinding spindle in %

10.  Cycle time tc

The person responsible for the grinding tests should always be able to track parameter changes across the testing timeline.

Publishing

It makes sense and is worthwhile to publish grinding results internally and to maintain a database of these results. For this reason, future processes can be designed, implemented, and optimized much faster. Sharing knowledge of good results - but also of failures -flattens everyone’s learning curve.

Part 2 of 4 parts will appear shortly

Walter Graf, The Philosopher’s Grindstones, August 2020