machining witness or reference point explanation;

[eyec]

(not for CNC programing, i have already read up of zero reference points - tool, machine & part) but to establish a beginning point in relationship to the whole part?
i am not a machinist but have seen some machining operations and heard machinists use the term witness point or reference point when setting up to machine a rather large part.
my question is (and i have exhausted many searches) where can i find it referenced in an industry standard or MSP.
tia for any assistance.

 

[MikeHalloran]

I don't think you'll find a standard explanation.

I will point out that it's often possible to fixture a part in such a way that when you come to some sequences in the program, there's no material to cut. In that case, a machinist will 'prethink' if you will, what the program does, so as to have some stock to cut for all the operations to be performed. A machinist might then refer to locating a 'reference point' (some arbitrary, but not really arbitrary) point or surface relative to the fixture so as to achieve the stated goal.

The term 'witness' or 'witness point' or 'witness mark' usually refers to a mark left on the workpiece, generally not by the active tool, but by a fixture or something else, perhaps in a previous operation, perhaps as part of some fitting sequence. Prussian Blue is sometimes involved in fitting operations. It's a dye that doesn't dry like Dykem. A little bit is applied to something, it's test-fitted in place, and sometimes a little affection is applied with a hammer. The blue is squeezed out where there was contact, leaving a 'witness mark', which may be used to locate something in the next operation, or used to guide setup of a cutter, e.g. to remove the high spot that made the mark.






Mike Halloran
Pembroke Pines, FL, USA

 

[eyec]

Mike,
thanks for the response. what you have described is what i have observed over the years.

my quandry is, i was looking at a skid that has some pedestals welded on it and the tops had to be machined "flat & true" to the base of the skid (within .001"). however, the drawing nor the skid baseplate identified a witness or reference point and no place on the baseplate of the skid was machined.

the skid is roughly 24'X12' with the pedestal tops @ 48". the machine shop laid the skid on its side to machine the pedestal tops!

so, how can i confirm the pedestal tops are true & flat to ensure the equipment that is going to be mounted on it will be level? each pedestal is true & flat in relation to each other but i cannot seem to find a way to confirm the basic premise that they will be so when the skid is set in place.

i was hoping to find a document that described what you said or i am used to so i can present it to the engineer and ask him how he can confirm the equipment will be level when mounted on the pedestals.

 

[MikeHalloran]

In all probability the skid was built on a flat floor, less likely on a platen, but in either case .001" tolerance is ludicrous. You are right; in order to establish a 'base plane' you would have to identify three points or near-points that would define a plane. Or just set it down on a floor and let the concrete do it. If you set it down in a thin layer of paint or mud, on a known flat floor, then you could infer where the real base plane is. Maybe that's what the machinist did.

Or maybe he just assumed the .001" tolerance came from a neophyte so he could ignore it and just do something reasonable, like coplanarity of .005" and parallelism to an unmachined base within 1/8" or so.

It probably doesn't matter anyway. Skids are normally jacked until the tops are level, and then the base is grouted. ... at which point it doesn't matter how flat the base is/was.



Mike Halloran
Pembroke Pines, FL, USA

 

[mfgenggear]

a ref plane will be established by machining the opposite side flat, the the part can be rotated & machined located on the the machine surface on the machine bed with or with out a tooling fixture.

thus the part is machined flat & parallel to a tooling machined flat surface.
thus the part can be inspected & recorded.
flatness can be inspect to itself like mike said. by locating on three place's then
verifying flatness by either an indicator or CMM .

a stand can either be verified to be level with a precision level ".0005" or better, or by theodolite"
which can inspect to .001" or better for lengths & that it is level.

HTH
Mfgenggear

 

[eyec]

mfgenggear, no place on the bottom was machined each skid weighs 16tons, so flipping the skid becomes a problem.

mikehalloran, to complicate the issue, i have 6 pedestals @ different heights (3 pairs)and piping connections that have to match up to existing piping. so leveling to such tight tolerances may be ludicrous after leveling the baseplate,but the drawing calls out the tolerance.

regardless, i would still like to find some document to reference to the engineer.

thanks for your responses.

 

[mfgenggear]

eyec

I was assuming you had a overhead crane?

Based on the lack of information the above was given. however it is not uncommon like Mike said, to jack up & shim the bottom of a machine bed.
to level the top. it's done every day by machinery movers. this where a level or theodolite is used depending on the level of accuracy.
there are target that are place on each corner or end of the machine top. from these target the theodolite measurements are given to ensure the level of accuracy is acceptable.
Major fixture Installations are also set up this way. I have been out of the loop so there may be major improvements like laser which is even more accurate.

there should be Information out there by google it?

HTH
Mfgenggear

 

[eyec]

Mfgenggear, i agree and understand the machining methods you mention, but i have performed many searches and not found a document to reference. i guess i will put it on the engineer and machine shop to answer how they can assure planarity prior to placing the equipment onto the pedestals.

thanks again

 

[mfgenggear]

EYEC

see this link

http://www.hamarlaser.com/systems/L740/index.htm

mfgenggear

 

[mfgenggear]

some more info

http://www.on-trak.com/laser-alignment.html

Mfgenggear

 

[dicer]

24 x 12 feet? I doubt there is a machine that can cut anything that size to that tolerance.
And besides if it is cut to that the difficult part is the realinement when it is taken off the machines table.
The simple act of lifting it will distort it like crazy. You will need a specialty alinement outfit to aline it, and likely they will also need to aline it for the machine shop as well. It may need to be finished when it is mounted in place where ever it is to be used. It may need to be hand scraped and or lapped into alinement. What will it sit on? 30 foot thick reinforced concrete slab?

 

[BillPSU]

When the machinist clamped the part to the machine they distorted the part more than the tolerance you are saying. Just changing the orientation from vertical to horizontal distorted the part.

This part will have to be shot in with lasers or optics when it is installed to even attempt to come close to the spec. The part will need adjusters to pull and push this part to the correct position. The part will also have to be mounted to a massive foundation to maintain the accuracy.

Now the other real world problem will be temperature. Machine the part at 70F and then install the part at 85F. Assuming the part is steel 24' section will grow about .030" just because of thermal expansion. Now load the skid with whatever mounts to it and watch the compression deflection.

The tolerances are unrealistic and not appropriate for this type of part.

Bill

 

[mfgenggear]

world largest milling machine.

http://americanmachinist.com/beyond...mpany-installs-worlds-largest-milling-machine

I once worked for General Dynamics this kind of stuff. it was done ever day.
the tooling required for major installations was held with in .001"

Once upon a time it had the largest machine shop & fabrication in the west coast of US.
if can be calculated it can be built. if can be built it can be calculated

HTH
Mfgenggear