Creating Simple Injection Molds

Hello all,

Complete newbie here. I am trying to create a simple aluminum injection mold using my NOMAD 3. All goes well according to plan until I start machining the second half of the mold. The second half ALWAYS seem to be off on the X axis and seems to be always around 0.03 to 0.05 inch off to the right. The Y axis is good from what I can tell.

Needless to say, the hole doesn’t line up and will not produce the kind of results I am expecting.

The 2 halves of aluminum are “identical” in size. They were purchased from a vendor who sells aluminum blanks that are “squared off” for small benchtop injection machines, like mine.

The NOMAD mills right through both sides with ease and does not seem to have any issues at all, so I don’t think anything is slipping, moving, or anything like that. The cavity is only 0.25 inches at the deepest.

On the software side, I am using Carbide 3d MESHCAM license (don’t know if that makes any difference), but I am able to generate the toolpaths and gcode with no problems.

I’m not so sure what I am missing here. The only thing I can think of at this point is that I am messing up with how I zero the machine?

Because of time constraints, I can only mill one of the 2 halves in a day. Then, I mill the second half a day later. So I zero the machine for the first half using BitZero V2, then all goes well, then I turn everything off and mill the second half the next day.

When I mill the second half, I zero the machine again using second half of the stock. Is this the correct way of doing this?

Any tips or pointers would be greatly appreciated. Thanks all.

Hi @RJE121908,

Welcome to the community.
Just checking: there is no 2-sided job involved, you mill each half of the mold separately from two pieces of stock, correct ?

Are the two halves of the mold identical in the CAD model ? (sometimes things are as simple as forgetting to flip the vector design to create toolpaths for the second half). Are you running the same toolpath for the first and second half of the mold ? You may want to double-check the generated G-code in a visualization tool (e.g. ncviewer.com which lets you navigate through the toolpath and check coordinates at specific points)

It could be the BitZero slipping during the probing operation, watch out for this (and you may want to have one finger on the probe during probing, just in case). You can and should also manually double-check X0/Y0 after probing, a 0.05" offset from where it’s supposed to be will quite be visible.

Can you post a picture of the two halves to illustrate where the offsets from stock border are off?

Since you mill two identical pieces of stock, you may want to install a jig (corner) to align the stock on the wasteboard in a specific location, zero off the lower left corner of the first piece of stock, run the first half, then run the second half without re-zeroing, just swapping the stock, the jig will guarantee it’s in the exact same location, which will remove the re-zeroing as a source of errors.

2 Likes

This begs the troubleshooting question, “Are you flipping the part on the X axis and Y movement is not changed?”

If the offset is in the X axis only (and that is the flip direction), the logic says that the toolpath for each side is not exactly centered to the flip; they may be mirrored, but they are different.

1 Like

As far as I understood, there is actually no flip involved at all, just the milling of two halves of a mold from two independent pieces of stock. I initially also read it as a 2-sided job, but it does not seem to be the case.

Yes, of course. I meant to convey, poorly, that the “flip” occurred in the software vectors.

2 Likes

Is he predrilling these for dowel/alignment pins also? or does he just just square them up? I assume you have some sort of alignment system baked into your mold model.

2 Likes

Thanks very much for the speedy reply Julien.

To answer your questions:

1.) NO, this is NOT a 2 sided job. I mill each half of the mold separately from 2 pieces of stock that are identical in size.

2.) When you say, Are the two halves of the mold identical in the CAD model ?, are you saying are they the same in regards to the size? If so, yes. Obviously, SIDE 2 will have the mirror effect, but the sprue hole should still line up. And they line up perfect when I view the files in Tinkercad. The models are designed by a 3rd party designer and he gave me a model for each mold half.

3.) When using MESHCAM, I use the exact same rates and parameters for each tool path.

4.) I started to lean more in the direction of BitZero could be slipping by a hair when I probe it. Can you walk me through how exactly I go about getting the X0/Y0 location after probing? Would that be in Carbide Motion some where? I am still trying to familiarize myself with the software.

5.) I will try to take a photo of it today and post it up to show.

6.) I like the jig idea! I think that’s brilliant! Is there a tutorial somewhere that can show I go about this setup? I have the low profile vise for the Nomad, but I am willing to buy the extra equipment if I have to. Also, this would mean that I shouldn’t turn off the machine, and just leave it on over night then until I get to SIDE 2 the next day? I hope I don’t get a power outage. =)

2 Likes

Oh, I just meant manually jogging to X=0, Y=0, from CM jog interface, and jog Z close to stock surface, to double-check that the tip of the endmill is still “visually” where it should be, not 1 or 2mm off. Of course if the zero is just a hair off you won’t detect it visually, but if it is 0.05" off you will definitely notice.

The only specific support for that action you have in CM is to button to go to [X0;Y0;Z0+6mm] in the jog panel

Actual Nomaders here will be in a better position to answer, but I often just mill a jig in some leftover thick MDF (mill a pocket slightly larger than your stock, with a dogbone in the lower left corner to make sure the (usually) square stock can fit snugly in the corner)

Not necessarily. The zero reference is persistent, so you should be able to power off, power on, re-home the machine, and return to the latest set zero. There is some (small) variability in the re-positioning accurary between one machine and the next, but it’s usually small enough that it’s acceptable/negligible.
If the re-homing/repositioning accuracy turned out not to be acceptable, another strategy is to manually jog the endmill to a known/fixed location on the jig (typically, a hole the diameter of the endmill), and (re)set zeroes there

1 Like

Oh ok, I see. So probe it, get coordinates for XYZ, and then jog the endmill to the corner of the stock (where I probed it with BitZero v2) and see if the coordinates are the same? Or at least close to the same? Did I understand you there correctly?

Oh, I thought you always had to zero the machine once started? I wasn’t aware that the location of “zero” stayed there even after being turned off.

Well, I’m going to try to make a jig as you suggested. I think it’s still a great idea and should keep the 2 halves aligned pretty well.

Since you have the low-profile vise, here’s my recommendation:

  • Square the vise with the machine (line up the 6mm dowel pin holes in the ends of the vise and the matching holes in the Nomad bed by putting a dowel pin through them before tightening down the screws)
  • Insert the stock
  • Zero using the Bitsetter
  • Mill your first piece
  • Remove it by loosening the moving jaw of the vise
  • Insert your next piece of stock but don’t tighten it yet
  • Put a dowel pin of known and precise diameter in the collet, or leave the endmill shaft poking out just slightly
  • Move the machine to Y minus the radius of whatever is in your collet, Z so the collet nut is nearly but not quite touching your stock, X somewhere in the middle of your stock (not important)
  • Push slide the stock in the vise towards whatever is in your collet until it touches
  • Fasten the vise

This should be reproducible because:

  • The X location of the stock should be the fixed jaw of the vise, which shouldn’t move between setups
  • The Z-location of the stock shouldn’t change
  • The Y location of the stock is set relative to the position the machine is already homed to

Also in hindsight, the vise not being square might be the reason for your problem in the first place. It’s important to line it up using dowel pins to get it squared properly (which for some reason C3D doesn’t provide). If you just bolt it down to the table it won’t be accurate.

3 Likes

When I machined some molds for pouring Jig heads around fish hooks, I didn’t rely on the edges of the aluminum stock to align the mold halves. Since those molds were hinged, I machined 3/8" hemispheres in both halves, and glued a 3/8" ball bearing in one half to align the mold halves.

1 Like

Didn’t even think about the vise not being squared! I’ll try this suggestion today and see what happens. Thanks very much.

1 Like

The 2 halves are already pre-drilled and have the pins installed when I get them. They are already “squared off” and aligns perfectly when put together. Thanks.

1 Like

Exactly.

It’s quite handy Indeed. It works out well if the machine has good homing repeatability (which is usually the case, but not if you are chasing micrometers…)

1 Like

I’m curious. What’s so special about the blocks you’re buying? They’re blocked out and dowel pinned for your injection machine, Isn’t that something you could do yourself when you get the hang of things?

I can’t thank you and Moded1952 enough! Thank you both so much for the help. It aligns perfectly now and I couldn’t have done it without you guys.

It turned out that the bitzero WAS slipping just a tad bit on the second probe operation on the X-axis and so I kept my finger on its side during the entire thing and when I jogged the mill back to the corner, it was nearly identical.

I also ended using up the fixed end of my low profile vise as the “jig” and simply butted up a corner of my stock with it, then when I got to the second half, I did not rezero and just ran the jobs. As Julien mentioned, the zero location stayed put and all was good =)

Thank you both very much!

1 Like

Thanks very much! The advice on the “fixed” end of the vise worked great and both halves line up perfectly now =)

Thanks to you and Julien. Thanks again and take care.

2 Likes

Nothing special about them Steve, except for maybe the fact that they cost $100 per set… =) I got a few of these because when I bought the injection machine, budget left me some room to purchase some add-ons. So, I figure I would rather just get a few of these so I can get to milling right away… But yeah, as I get more comfortable with the NOMAD, I am sure I can just drill the holes myself and get the pins from mcmaster or something like that and press the pins in. I don’t know how difficult that is, but I am sure it can be done. I’ve read from other forums that folks do the exact same thing.

Take care.

1 Like

Sort of on topic, what kind of injection molding machine do you have?

1 Like

PIMS 150A from LNS Technology

3 Likes