Harig 612 Reassembly - Table Flatness Measurement

[jwweathers]

I am reassembling my Harig 612, and I measured the table flatness with 10ths dial indicator (see picture below) from the column. Any thoughts on what to do about this sort of wear?

Thanks!
Jeff

 

[eKretz]

We're going to need a little better description of what and how you're measuring. A photo or two showing your measurements in progress would be helpful. If you are just running an indicator on the table with table travel, that's not actually going to tell you a whole lot. A better quick test is to set a small granite surface plate of known provenance on the table and indicate that - then at least you know you're indicating a very near-planar surface.

 

[jwweathers]

We're going to need a little better description of what and how you're measuring. A photo or two showing your measurements in progress would be helpful. If you are just running an indicator on the table with table travel, that's not actually going to tell you a whole lot. A better quick test is to set a small granite surface plate of known provenance on the table and indicate that - then at least you know you're indicating a very near-planar surface.

I attached a 10ths dial indicator to the column of the grinder using a magnetic base. I zeroed the indicator in the front left corner of the table. I mapped the table surface with the indicator assuming that would give me some idea of its flatness relative to the spindle. I will post a picture or two tomorrow morning.

 

[eKretz]

I attached a 10ths dial indicator to the column of the grinder using a magnetic base. I zeroed the indicator in the front left corner of the table. I mapped the table surface with the indicator assuming that would give me some idea of its flatness relative to the spindle. I will post a picture or two tomorrow morning.

Nope, won't give you any real indication of flatness unless you know the surface you're indicating is dead flat. An indicator reading true (or not) to an unchecked surface does not necessarily indicate that the surface is (or isn't) flat. You need a reference flat or a way to check the surface you are indicating against one to get an actual idea of whether the machine table is traveling in a straight/planar manner.

 

[jwweathers]

Nope, won't give you any real indication of flatness unless you know the surface you're indicating is dead flat. An indicator reading true (or not) to an unchecked surface does not necessarily indicate that the surface is (or isn't) flat. You need a reference flat or a way to check the surface you are indicating against one to get an actual idea of whether the machine table is traveling in a straight/planar manner.

Makes sense. I suppose the table could travel in an arc which could indicate true with the method I am using. With there being some difference in the measurements in my case, there is some error somewhere. It is just a question of the source of the error (e.g. table travel or flatness). Am I thinking about it correctly?

Thanks!
Jeff

 

[RJT]

Are you measuring the surface of the chuck after grinding it, or the top of the saddle that the chuck attaches to?

 

[jwweathers]

Are you measuring the surface of the chuck after grinding it, or the top of the saddle that the chuck attaches to?

I am measuring the surface of the saddle that the chuck will sit on. I am calling it the table. May be the wrong term…

Thanks!
Jeff

 

[eKretz]

Makes sense. I suppose the table could travel in an arc which could indicate true with the method I am using. With there being some difference in the measurements in my case, there is some error somewhere. It is just a question of the source of the error (e.g. table travel or flatness). Am I thinking about it correctly?

Thanks!
Jeff

Yes. Your indicated error is probably from a combination of those issues. That's why using a known flat reference is helpful - that lets you determine exactly where the error actually is - in the table surface itself or in the ways.

 

[david n]

Most likely not even an issue..............throw the mag chuck on and grind it flat....................

 

[jwweathers]

Most likely not even an issue..............throw the mag chuck on and grind it flat....................

My kind of plan!

 

[eKretz]

It never hurts to know what you're working with... Checking it out doesn't mean you need to get right on with fixing it if there's even a problem. Personally, I figure it's way better to know what you're working with so you can work around problems if/when necessary. If you're just using the grinder to make things look shiny, that's another story.

 

[MCritchley]

I would call your measurements “sweeping the table”
Those numbers are not bad, it’s not a new Mitsui. Grind the chuck and make some parts.

Was this the machine with a broken cross feed screw? You ended up with a good machine that only has .0004” variance from a previously ground table to the spindle.
Again this variance means very little.

 

[jwweathers]

It never hurts to know what you're working with... Checking it out doesn't mean you need to get right on with fixing it if there's even a problem. Personally, I figure it's way better to know what you're working with so you can work around problems if/when necessary. If you're just using the grinder to make things look shiny, that's another story.

I completely agree. I was being a bit facetious with the “my kind of plan” comment. Although, I am anxious to start making some parts.

Thanks!
Jeff

 

[jwweathers]

I would call your measurements “sweeping the table”
Those numbers are not bad, it’s not a new Mitsui. Grind the chuck and make some parts.

Was this the machine with a broken cross feed screw? You ended up with a good machine that only has .0004” variance from a previously ground table to the spindle.
Again this variance means very little.

This is that machine. I disassembled it completely, cleaned, painted, and I am in the process of reassembling it. The table sweep is the first check I have made. I am excited to get it up and running. I have a 9” camel back straight edge I want to grind.

Thanks!
Jeff

 

[michiganbuck]

The chuck top can have wear so showing an error due to chuck condition. Traveling an indicator over a fresh ground chuck getting zero tenths indicator trave check end to end and on the cross, a part ground there can have a .003 (or what) concave or convex on long or cross.. To be really zero the long and cross ways need to be dead flat and then the chuck ground.

A good simple test is to precision straight edge the top of a chuck close to the operator's position just to give a rough idea if the last chuck grinding made a flat chuck.

 

[jwweathers]

The chuck top can have wear so showing an error due to chick condition. Traveling an indicator over a fresh ground chuck getting zero indicator trave check end to end and on the cross, a part ground there can have a .003 concave or convex on long or cross.. To be really zero the ways need to be dead flat and then the chuck ground.

Right. I suppose one way to check that is to grind a part and inspect it on a surface plate, correct?

 

[eKretz]

Right. I suppose one way to check that is to grind a part and inspect it on a surface plate, correct?

Yep. However, you've got to watch out for the magnet pulling a part down or heat warping it if you go that route. Lots of pitfalls to be aware of.

 

[michiganbuck]

The long-travel bed ways can be out of wack and so the top grinding of the chuck may indicate good off the wheel head, perhaps zero after grinding the chuck and part grinding is not flat. the problem here is that the chuck top may be concave or convex after top grinding, reflecting the condition of the ways.
Such poor long bed flatness can be the result of lack of oil, air hose use near the machine, over-tightening the chuck. too heavy a chuck, prolong production job stopping at the same place reveres, ball ways worn, and likely other reasons.
Yes, sometimes this can be on the cross.
Grinding a part on an out-of-flat chuck can give a lot of confusing results, and can vary with the size of the part or where it is set on the chuck..

A simple test is to check that the chuck top is near dead flat after grinding the
chuck. or shim up a dead flat parallel on the chuck zero/zero and run a tenths indicator on top of that before grinding the chuck.
A rough test is to put three or four same height gauges on the chuck long ways and touch them noting that the down dial marks the same number (Might check for 1 to 3 tenths).

 

[jwweathers]

The long-travel bed ways can be out of wack and so the top grinding of the chuck may indicate good off the wheel head, perhaps zero after grinding the chuck and part grinding is not flat. the problem here is that the chuck top may be concave or convex after top grinding, reflecting the condition of the ways.
Such poor long bed flatness can be the result of lack of oil, air hose use near the machine, over-tightening the chuck. too heavy a chuck, prolong production job stopping at the same place reveres, ball ways worn, and likely other reasons.
Yes, sometimes this can be on the cross.
Grinding a part on an out-of-flat chuck can give a lot of confusing results, and can vary with the size of the part or where it is set on the chuck..

A simple test is to check that the chuck top is near dead flat after grinding the
chuck. or shim up a dead flat parallel on the chuck zero/zero and run a tenths indicator on top of that before grinding the chuck.
A rough test is to put three or four same height gauges on the chuck long ways and touch them noting that the down dial marks the same number (Might check for 1 to 3 tenths).

I think I get it. Do something to eliminate the table flatness as a variable to isolate the source of the error.

Seems like your last suggestion using the gauges is the same as sweeping the table like I’ve already done, but I may be missing something.

Really appreciate the suggestions. My immediate plan is to get the machine running and see what the parts look like.

I have included a few pictures of where I am in the reassembly for anyone interested.

Thanks!
Jeff

 

[michiganbuck]

Sweeping the table by indicating the chuck top off the headstock ( wheel spindle head) might check/show zero-zero but that does not prove the machine will grind a flat part. Checking for bed way condition only a flatness test proves that the fresh ground chuck is truly flat and assures that grinder will grind a long pat flat. *It is different than sweeping the chuck and different from the set gauge blocks on the chuck.
If the chuck after grinding has a slight concave or convex top then sweeping may check dead true but the chuck out of flatness, along with the out of flat travel will make a ground part out-of-flat.
It is good/best to inspect the set pad for flatness and check the chuck top for flatness not with sweeping but with a near dead true flatness gauge of some sort.
Note that I said the three or four gauge block test is a rough test/ the same as sweeping the chuck is a rough test.
I have seen some of the best name grinders with having this condition and it makes grinding high precision parts a confusing nightmare because some grinder hands have difficulty figuring out why part-error occurs, or occurs on some parts and not all parts
I guess every precision grind shop should have a precision flat something on
hand and use it with not turning on a strong magnet that might suck it out of flat.