Multi sided ops with 3 axis machine. Can I make this any easier?

For 13 years now I've been building injection molds with 3 axis vertical mills. When it comes to making inserts there is always multi sided work that needs to be done. We have always tackled it by doing the back work first, inserting taped and dowel holes where possible and then attaching a fixture plate to it so we can move the fixture plate, and attached insert, to an angle plate or 246 block to manually rotate and do the side work.

Business is changing a little and we are seeing more small batch prototype work that needs the same multi sided ops to complete the parts. Most of these can't have the taped and doweled holes for fixturing so it's a lot of set-ups in the vise to hit all sides. Is there an easier way to approach this than what we are doing currently?

I've been looking into the Fifth Axis self centering vises and their accompanying 90 degree angle plates. Thinking that I could set up the vise on the table to do the top/bottom work then move it to the 90 degree angle plate to rotate through the 4 sides. My fear is that unless I get the vise base indicated in PERFECTLY on the table, I'm going to be off on my side work. Not that big of a deal when playing with +/- .005 for the parts work but when mold inserts have to be held within .0001's, I have my doubts.

I don't have the real estate on my table to dedicate to a 4th axis due to large mold bases going in and out all the time.

How are others approaching this and what might I be missing here?

Thanks.

Ed

We have 3 of the 5th axis Vise’s 1 6 and 2 5 s they don’t repeat well enough and if your part isn’t exactly centered up the vise twist too easily we probe and align the c axis every time to account for it. I would try another brand myself. Maybe someone else will chime in with more help
Don


Sent from my iPhone using Tapatalk Pro

YMMV, but I like the Lang stuff. I'd probe in on the 1st op "close enough", cut it, then put it in op2 and probe off of op1 features.

RCS Machine said:

How are others approaching this and what might I be missing here?

Click to expand...

How about a horizontal ?

EmanuelGoldstein said:

How about a horizontal ?

Click to expand...

Hmc 500?

Sorry couldn't help it.

plastikdreams said:

Hmc 500?

Sorry couldn't help it.

Click to expand...

You mean the one empwoer is having trouoble with ? Sounds like he'd sell it cheap

EmanuelGoldstein said:

You mean the one empwoer is having trouoble with ? Sounds like he'd sell it cheap

Click to expand...

its not mine, but it is for sale

Hi RCS Machine:
If you only do orthogonal 3 +2 axis type of work, you could mount your mold inserts on a little pallet, and just manually flip your inserts using a mounting fixture.
3R makes something like this for mounting EDM electrodes and they are very accurate...well within the range for your needs.
It's a bit like having a 5 axis machine except you have to have to do the reorient manually and obviously you can't do 4 or 5 axis interpolation, but for only orthogonal 3+2
It's a technologically backward but completely viable way forward.

3R makes pallets for production machining with dovetail vises built in.
Now you'd only need to bolt the indexing block onto your table, or better yet, mount up two...one face up and one on it's side.
If you dial them in perfectly, and mount them at one end of the table so you can still get a mold base on the machine without ripping them off again, you can whale away any time on 5 sides of a block, take it off, inspect it on the CMM and remount it within a tenth in seconds if you want to.
You can mount another pair on a surface grinder, and a third one on a wire EDM and leave them set up there too.

I've never personally felt the need for something like this, but if your circumstances are different from mine, maybe you do feel the need.
I don't know how hard you can push them...I've only ever used the system for making electrodes, but they do advertise the capability for milling, so I'd certainly contact 3R and ask them.
I do know they're super expensive.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

I run a Lang setup with one quick point plate flat on the table and another on an angle block.
I took great care to dial in the plates and angle blocks, setting each one on its own WCS.
I believe the lang plates are repeatable to 8 microns.

For my purposes it work very well, I program all my parts from the center top of the plates, clamp in the stock and go.
I have found on my machine the largest source of error is the Thermal growth in the Z axis.

For the sort of tolerances you are working to this system probably isn't accurate enough though.
I can get bores inline concentric on opposite sides of a part within 20microns, that's about as good as I seem to get.

implmex:

When I started toying with this over a year ago, I looked into the 3R system but wrote it off for some reason. Time flies and I can't remember why now but the manual flipping of sides is exactly what I was looking at doing. I've used them for cutting trodes in this fashion so that's where the thought of inserts came from.

srp61:

This is what I was thinking of doing with the Fifth Axis system. Have one base plate mounted to the table and do my 1st op work in the vise then take the vise off the base, replace it with their angle plate and put the vise on that. The goal is that once the workpiece is in the vise I want it to remain in the same vise until darn near finished to reduce any re-clamping errors.

It seems that I have read more positive reviews on the Lang system than others out there. I have a large cube of 7075 to make a tombstone from to mount the quick point plate on. I can have my buddy cut my cube faces in his 5 axis machine to get the sides perpendicular to each other and just take the time to dial the mounting plates in so they match each other. My only holdup is the unknown accuracy I'll find when moving the vise from the table mounting plate to the plate on the tombstone since there is no way to rotate things since they are all static at this point.

Hi again RCS Machine:
I think we are all in agreement that clamping the workpiece onto a pallet is the way forward for you, so long as all the pallets in your pool are very close to identical and so long as your indexing block is very accurately square and parallel and you've dialed it in as accurately as you and the machine are capable of.
What system it happens to be no longer matters so long as it meets those criteria and is robust enough to tolerate what you are going to do with it and is big enough to handle the largest parts you intend to machine this way.

If you cannot find something suitable and if your need can justify all the work, there is no theoretical reason you could not roll your own...if you build molds for a living you certainly have the skills and the equipment too.

Since you get to control the dimensions of the dovetails you prep onto the blocks, I'd dispense with the fancy, unreliable, inaccurate, and flimsy self centering vises, and just make mounts like the pictures below so you can attach your parts using the same dovetail dimensions every time.

It's no hardship to do so, and you get the benefit of having them all the same and get much better repeatability too, if you ever need to take the block off it's pallet for some reason and then re-mount it.

I wire cut all my dovetails because it's easy to do so and utilizes a machine that would otherwise often sit idle, but there is no reason on earth why you MUST do so too.
My stuff is made so I can hang a titanium bar off the end of the platen in the wire EDM worktank so there's no need for me to attach it to a pallet.
What I've shown is the remnant of a 2" titanium bar after I've wirecut the blanks for a freaky sex toy from it.

My fixtures were just knocked together from scraps of 316 SS round bar I happened to have kicking around, so they look pretty home made.
Obviously my system works a little differently from how your design would work, but that's just a detail.

So if the commercial offerings are not to your taste, these are not hard fixtures to make and they do work very well.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

RCS Machine said:

srp61:

This is what I was thinking of doing with the Fifth Axis system. Have one base plate mounted to the table and do my 1st op work in the vise then take the vise off the base, replace it with their angle plate and put the vise on that. The goal is that once the workpiece is in the vise I want it to remain in the same vise until darn near finished to reduce any re-clamping errors.

It seems that I have read more positive reviews on the Lang system than others out there. I have a large cube of 7075 to make a tombstone from to mount the quick point plate on. I can have my buddy cut my cube faces in his 5 axis machine to get the sides perpendicular to each other and just take the time to dial the mounting plates in so they match each other. My only holdup is the unknown accuracy I'll find when moving the vise from the table mounting plate to the plate on the tombstone since there is no way to rotate things since they are all static at this point.

Click to expand...

So I surface ground my block square, Even if you get your block machined on a 5 Axis mill id still verify that its actually square, not all 5 axis machines are that accurate,
Also you need to verify the squareness to the machines travels one its installed, I have mine dialed to within 2 microns.

The hard part is setting the WCS accurately, I originally didn't have the Lang Gauging pallet so I machined some features then rotated the vice arounds the 4 different orientations and recorded the coords, then used a circle calculation on the Heidenhain control to find the true center for the flat plate. the angle plate is a bit trickier, finding X/Y is fairly straightforward, Z is a moving target due to my machines thermal growth. I have found an acceptable solution is to freshly calibrate my probe and tool setter, then set the Z for the angle block WCS, Then I have to use the tool probe to probe each tool after I call it up to update the length as the temperature changes during the day.

The machine has Z thermal comp but it doesn't work very well and we cant temperature control our 3000sqM workshop.

TLDR, Setting the WCS and plate alignment is critical to it working well

Also Id never use 5th axis stuff, just seems too flimsy, too much is made from aluminium, Lang stuff is super nice and incredibly accurate.

srp61 said:

So I surface ground my block square, Even if you get your block machined on a 5 Axis mill id still verify that its actually square, not all 5 axis machines are that accurate,
Also you need to verify the squareness to the machines travels one its installed, I have mine dialed to within 2 microns.

The hard part is setting the WCS accurately, I originally didn't have the Lang Gauging pallet so I machined some features then rotated the vice arounds the 4 different orientations and recorded the coords, then used a circle calculation on the Heidenhain control to find the true center for the flat plate. the angle plate is a bit trickier, finding X/Y is fairly straightforward, Z is a moving target due to my machines thermal growth. I have found an acceptable solution is to freshly calibrate my probe and tool setter, then set the Z for the angle block WCS, Then I have to use the tool probe to probe each tool after I call it up to update the length as the temperature changes during the day.

The machine has Z thermal comp but it doesn't work very well and we cant temperature control our 3000sqM workshop.

TLDR, Setting the WCS and plate alignment is critical to it working well

Also Id never use 5th axis stuff, just seems too flimsy, too much is made from aluminium, Lang stuff is super nice and incredibly accurate.

Click to expand...

I'm not dissing 5th axis "stuff",

But I have to say that is one of the very tempting things about a Makino [3 axis , boxed way ultra precise machine.] V33i and similar is that the whole build geometry and thermal control is within a few microns. like less than 5 micron (not 5 tenths) + really good thermal growth Z control with pretty incredible spindles (for what they are designed to do.). [Insulated thick castings so thermal changes in a shop are mitigated to a large degree + scales (of course)].

The machine can make/ cut measured larger circular interpolated features to 1 micron to 1.5 micron. The table geometry (flatness) is within a couple of micron, sub micron straightness of travels, steep walls cut - flat - to sub 3 micron... and on and on,

The machine is practically a CMM and optioned out a certain way - not just having jig boring accuracies , you can actually perform various (wet) grinding operations + jig grinding cycles. [+ switchable option for machining graphite for sinker EDM applications.].

So with inspection level angle plates and other precision (static and dynamic) fixtures like air spindle rotary bearings etc. there's a lot of high quality work that can done with one super accurate 3 axis machine. [more hands on in "mind-set".]. ~ but slower work.

For smaller mold cores seems Makino are keen to push the 5 axis - D-200 Z (5 axis accurate machine) - for parts in a pretty wide range of materials - work volume pretty much the size of a medium to large size (taller) coffee can. But the base price for a D-200 Z is about $370K

Whereas something like V33i (3 axis machine) (new) starts around $270K. Makino do make a +/- 1 arc second-ish small 5 axis trunnion unit / rotary but that does take up a lot of good real-estate (table wise) on a V33i or even V56i

The Makino F3 and F5 are more of a compromise being built in Singapore (all good) but with linear rolling element slides rather than box ways of extreme precision. $175 to $250 K proposition but probably more 'Aluminum" / Aluminium friendly.


_____________________________________



V33i and V56i It's like a reference surface + cmm + jig grinding, jig boring and mold capable mill in one + very accurate mechanical and optical mold making "machine". + very stable over long periods of time + SGI ("Super Geometric Intelligence" - at least great / sensible / smart / clean surfaces and contours.).

I know that's a little "Off topic" vis a vis precision rotatable / indexable and referenceable clamping and fixture systems + thermal management / z control + Mold core issues.

Time spent correcting 5 axis machine work on three or four other machines vs. multiple set ups on orthogonally ultra precise and stable 3 axis equipment ? + compatible clamping and fixture systems + custom made fixtures/ fixture plates in-situ. + clever things that can create supports and reference surfaces as part of the part geometry to be machined off later or use of tooling balls etc.<_--- that capability comes about from being able to re-reference parts accurately through the day / thermal temperature changes + super excellent build geometry. "Regular" 5 axis machine processes basic claim to fame is fewer "Screw ups" between set-ups. Other than sim 5 axis geometry.

RCS Machine said:

For 13 years now I've been building injection molds with 3 axis vertical mills. When it comes to making inserts there is always multi sided work that needs to be done. We have always tackled it by doing the back work first, inserting taped and dowel holes where possible and then attaching a fixture plate to it so we can move the fixture plate, and attached insert, to an angle plate or 246 block to manually rotate and do the side work.

Business is changing a little and we are seeing more small batch prototype work that needs the same multi sided ops to complete the parts. Most of these can't have the taped and doweled holes for fixturing so it's a lot of set-ups in the vise to hit all sides. Is there an easier way to approach this than what we are doing currently?

I've been looking into the Fifth Axis self centering vises and their accompanying 90 degree angle plates. Thinking that I could set up the vise on the table to do the top/bottom work then move it to the 90 degree angle plate to rotate through the 4 sides. My fear is that unless I get the vise base indicated in PERFECTLY on the table, I'm going to be off on my side work. Not that big of a deal when playing with +/- .005 for the parts work but when mold inserts have to be held within .0001's, I have my doubts.

I don't have the real estate on my table to dedicate to a 4th axis due to large mold bases going in and out all the time.

How are others approaching this and what might I be missing here?

Thanks.

Ed

Click to expand...

Sorta ^^^ leaning into the original post / conundrum. I agree that the EDM "Peeps" have some good solutions in general for not having to buy additionally expensive DD or servo driven rotaries etc., or even lean on Erowa or 3R type pallet system fittings (or at least has been recommended to me in the past.).