- Joined
- Dec 29, 2012

- Messages
- 337

Some two years ago I was not happy with my 125mm 4jaw independent chuck performance. After cleaning the chuck (removing all jaws) I noticed that after centering a round bar, the indicator showed 0.0004" (0.01 mm) near the chuck jaws, but when I moved the indicator further away from the chuck to a distance of a few inches, the readings multiplied by a factor between 10 and 50!!!

This showed a very serious misalignment of either the chuck, the jaws or both of them.

Usually, when I needed to machine a long bar I always supported it to the tailstock so I never made any measurements at a distance from the chuck.

After some tests, while I put each of the 4 jaws in different positions and made measurements, I noticed that the readings, at the end of the bar, changed in a more or less random way according to where each jaw was fitted!!!.

There was no way to keep the near-zero indication at the end no matter the (random) jaw position!!!

To find the position with the smallest error I planned a more systematic measurement approach.

First I always bolted the chuck at the same position on the plate as you can see from the marks both at the chuck and the plate (in the middle of the chuck).

Next I named the positions on the chuck I, II, III and IV and the jaws 1, 2, 3 and 4.

I thought initially that the possible combinations between positions and jaws were 4X4= 16 but after a small research I found out that all possible combinations of the numbers 1, 2, 3 and 4 are calculated in a different way:

We will use each of the four numbers as the first number. For example number 1 will be the first.

With the rest of the jaws there are 3 combinations for the second number

1-2

1-3

1-4

In total (to use each number as the first): 4 numbers × 3 choices = 12 ways to choose the first two numbers.

After using the first two jaws (numbers) there are only 2 more jaws (numbers) left to choose as a third number. That's 4×3×2 or 24 ways to combine the first three jaws.

After the first 3 numbers there is only 1 way to choose the fourth. That's 4×3×2×1 or still 24 ways

All the possible combinations are as follows

1. 1234

2. 1243

3. 1324

4. 1342

5. 1423

6. 1432

7. 2134

8. 2143

9. 2314

10. 2341

11. 2413

12. 2431

13. 3124

14. 3142

15. 3214

16. 3241

17. 3412

18. 3421

19. 4123

20. 4132

21. 4213

22. 4231

23. 4312

24. 4321

So I had to make

In other words, I needed to make 24 changes of the 4 jaws plus 24 indications at each end of the bar and record the data... That

So I started making all possible combinations of the jaws to the chuck, cenered the bar as close to the chuck as possible and then make measurements of the bar at a few inches (5") distance.

As you can see from the above data The best choice was to use the combination with the smaller error and try to correct it further.

For that reason I indicated a test bar to the 4 jaw chuck, using the smaller error jaw combination and then without removing the bar I removed the chuck and supported by the bar in between centers.

Next I very lightly skimmed the back of the chuck.

After mounting the chuck on the lathe the error was still there but now was 0.002" (0,05 mm) 5" away from chuck, (at the same distance I made the measurements before).

I was so bored from the numerous jaw changes and measurements that I called it DONE.....

Further measurements needed but at least I was happier with it as it was at that time.

To avoid any future mix up of the jaws, I marked each position on the chuck and also marked each jaw using the same mark.

I wished then my next 4 jaw independent chuck to have 4 identical jaws!!!

This chuck kept me happy for another 2 years before it started misbehave again, but this is another story to be published soon, very soon to be precise!

Thank you for reading

Petros

This showed a very serious misalignment of either the chuck, the jaws or both of them.

Usually, when I needed to machine a long bar I always supported it to the tailstock so I never made any measurements at a distance from the chuck.

After some tests, while I put each of the 4 jaws in different positions and made measurements, I noticed that the readings, at the end of the bar, changed in a more or less random way according to where each jaw was fitted!!!.

There was no way to keep the near-zero indication at the end no matter the (random) jaw position!!!

To find the position with the smallest error I planned a more systematic measurement approach.

First I always bolted the chuck at the same position on the plate as you can see from the marks both at the chuck and the plate (in the middle of the chuck).

Next I named the positions on the chuck I, II, III and IV and the jaws 1, 2, 3 and 4.

I thought initially that the possible combinations between positions and jaws were 4X4= 16 but after a small research I found out that all possible combinations of the numbers 1, 2, 3 and 4 are calculated in a different way:

We will use each of the four numbers as the first number. For example number 1 will be the first.

With the rest of the jaws there are 3 combinations for the second number

1-2

1-3

1-4

In total (to use each number as the first): 4 numbers × 3 choices = 12 ways to choose the first two numbers.

After using the first two jaws (numbers) there are only 2 more jaws (numbers) left to choose as a third number. That's 4×3×2 or 24 ways to combine the first three jaws.

After the first 3 numbers there is only 1 way to choose the fourth. That's 4×3×2×1 or still 24 ways

All the possible combinations are as follows

1. 1234

2. 1243

3. 1324

4. 1342

5. 1423

6. 1432

7. 2134

8. 2143

9. 2314

10. 2341

11. 2413

12. 2431

13. 3124

14. 3142

15. 3214

16. 3241

17. 3412

18. 3421

19. 4123

20. 4132

21. 4213

22. 4231

23. 4312

24. 4321

So I had to make

**EACH ONE**of the above to cover all possible combinations!!!In other words, I needed to make 24 changes of the 4 jaws plus 24 indications at each end of the bar and record the data... That

**TOOK TIME**!!!So I started making all possible combinations of the jaws to the chuck, cenered the bar as close to the chuck as possible and then make measurements of the bar at a few inches (5") distance.

As you can see from the above data The best choice was to use the combination with the smaller error and try to correct it further.

For that reason I indicated a test bar to the 4 jaw chuck, using the smaller error jaw combination and then without removing the bar I removed the chuck and supported by the bar in between centers.

Next I very lightly skimmed the back of the chuck.

After mounting the chuck on the lathe the error was still there but now was 0.002" (0,05 mm) 5" away from chuck, (at the same distance I made the measurements before).

I was so bored from the numerous jaw changes and measurements that I called it DONE.....

Further measurements needed but at least I was happier with it as it was at that time.

To avoid any future mix up of the jaws, I marked each position on the chuck and also marked each jaw using the same mark.

I wished then my next 4 jaw independent chuck to have 4 identical jaws!!!

This chuck kept me happy for another 2 years before it started misbehave again, but this is another story to be published soon, very soon to be precise!

Thank you for reading

Petros

Last edited: