Router Forums banner
1 - 20 of 24 Posts

·
Registered
Joined
·
44 Posts
Discussion Starter · #1 ·
Any one use one?

Interested in pros and cons over modeling, I'm looking (again) at transitioning from my duplicator to CNC for a part I make over and over. Its going to be hard to model being full of weird curves.
 

·
Registered
Joined
·
198 Posts
i do not have one but have been interested. i know that a good cnc and a digitizer/probe attachment is all the hardware that is needed to be a decent digitizer. in my opinion, it will be the software package that is running the show which can make or brake how well it performs.

what i have read from users is that it is very slooooow. but, that may not be a deal breaker...

will be watching this one for sure!
 

·
Registered
Mike
Joined
·
3,836 Posts
The model being full of weird curves might be a problem but, as long as there are no undercuts, there should not be a problem. The thing to remember is the more detailed the model the longer it will take to get a good scan of the model. Also, the files will be quite large and will probably need to be edited so it might take quite a while before you can get a usable file to carve. One other problem is the size of the detail in the item. If there is a lot of very fine detail, most probes will not be able to reproduce that part of the item because of the size of the probe tip and the small stepovers required to follow the details. You might be able to overlay some detail on the model after it is created to resolve that problem but that is just one more thing to consider.

The good thing is that once you have probed the item and edited the files you should be able to mass-produce the item without having to be right there with the duplicator moving the tracing point.

One other thing to think about would be taking the item to a business that has a digital 3D scanner that can produce a model for you. I know this is expensive but the question is, "Can the price be justified for time saved and the quality of the model they would be able to produce for you?".
 

·
Registered
Joined
·
152 Posts
Okay, I am a newbie with just enough experience to be dangerous, so I'd like to learn something about these probes. I have a friend who was given his great-grandfather's shotgun. The issue is that the stock is broken in half right at the grip. He wants to keep as much of the stock as possible but the break wasn't clean and would need some repair. Part of the wood is missing so it just cannot be glued back together. He and I are thinking that maybe there is a method of tuning something to match on the CNC rotary axis, but how do you get the A) shape identical, and B) mill both parts to make the repair piece fit.

So, I'm reading about these probes and am completely unfamiliar with them. Would someone be kind enough to post a link or anything that describes what they are and how they work. Are they hugely expensive? Maybe Mike's mention of a 3D scanner is the solution where you could simply build a duplicate stock and cut out the part needed? It is not that important but I am just interested in learning more about the technology of both systems and learning more about how they would apply to such a situation.

I wasn't aware of this technology until I read your posts above. Thanks guys!

Mike
 

·
Registered
Joined
·
152 Posts
PS - Just thinking here, I am learning to repair some high-value antique wall regulator clocks with a friend. He's the expert at the clock mechanism, not me. However, when the clock has a broken fenial or part, a scanner could be excellent at digitizing the matching one, creating it on a rotary axis, and assuming that the stain can be matched, repair done without hours of adjusting the model in Aspire to get it perfect.

I looked at the two storehouses of knowledge, Google and Amazon and saw the small 3D scanners, both handheld and, for lack of a better term, turntable type scanners. Some are really cheap (in relative terms) at +/- $200 while there are others that are $20,000 +. Quite a range.

Another thought is that I have a friend who does independent work for companies where he sets up scanner(s) in industrial buildings and plants and maps the whole facility in a digital manner down to pipe sizes, etc. I know that has to be MEGA bucks for something like that, but if I could persuade him to help me, is that the same type of technology? I'm assuming it is close but WAY overkill.
 

·
Registered
Mike
Joined
·
3,836 Posts
I think I would look at repairing the broken stock if all the parts are there since it has sentimental value and will probably be passed down to future family members.

If part of it is missing and there is at least one good side then a digitizing probe could help create a model of one side then that could be mirrored to the other side to create the full model needed. the only problem I see is, as I said before, would be with the diameter of the probe tip. Any large curves force the probe away from the model, but with editing to the file this error can be lessened to produce an acceptable model.

One other thing to consider is to software used to run the probe with. If it is based on a flat(like most software are)surface then you with have those small errors. If it is based on using a rotary axis while probing them you should be able to get a very good solid 360 degree 3D model.

Just saw your second post.
If you have probing software base on a rotary axis then you could use it to scan a part like a finial or gun stock and reproduce it but remember this is a very time-consuming process. The question exists, "Is it worth the time to use the probe and edit the file for just one single piece?". If it is an item that will be made over and over then I would say, do it, it is worth the time involved. It would be worth digitizing a file for a shotgun stock if I wanted to reproduce that stock as a custom aftermarket replacement that would be relief carved with a choice of bird hunting scenes in a dish or special checkering. With that in mind, remember, you have to have a market for that type of item before spending all the time to get to that point.
 

·
Registered
Joined
·
152 Posts
I think I would look at repairing the broken stock if all the parts are there since it has sentimental value and will probably be passed down to future family members.

If part of it is missing and there is at least one good side then a digitizing probe could help create a model of one side then that could be mirrored to the other side to create the full model needed. the only problem I see is, as I said before, would be with the diameter of the probe tip. Any large curves force the probe away from the model, but with editing to the file this error can be lessened to produce an acceptable model.

One other thing to consider is to software used to run the probe with. If it is based on a flat(like most software are)surface then you with have those small errors. If it is based on using a rotary axis while probing them you should be able to get a very good solid 360 degree 3D model.

Just saw your second post.
If you have probing software base on a rotary axis then you could use it to scan a part like a finial or gun stock and reproduce it but remember this is a very time-consuming process. The question exists, "Is it worth the time to use the probe and edit the file for just one single piece?". If it is an item that will be made over and over then I would say, do it, it is worth the time involved. It would be worth digitizing a file for a shotgun stock if I wanted to reproduce that stock as a custom aftermarket replacement that would be relief carved with a choice of bird hunting scenes in a dish or special checkering. With that in mind, remember, you have to have a market for that type of item before spending all the time to get to that point.
Thanks Mike!
 

·
Registered
Joined
·
44 Posts
Discussion Starter · #8 ·
Another thought is that I have a friend who does independent work for companies where he sets up scanner(s) in industrial buildings and plants and maps the whole facility in a digital manner down to pipe sizes, etc. I know that has to be MEGA bucks for something like that, but if I could persuade him to help me, is that the same type of technology? I'm assuming it is close but WAY overkill.
I did have a part scanned once for production of a mold for duplicating the thing in EPDM rubber. Cost was $500 as I recall, and the part was relatively simple so there was probably a minimum of editing required. Probably a bit much for your situation, but still feasible for mine given Mike's comments (which were very helpful!).

I do know a bit about stock making if you want some input. The thing that many people find a little counter intuitive is that you usually do the inletting (fitting to the metal parts) first so that the flat planes of the stock blank can be used as reference surfaces. If I were making a replacement piece by hand I would approach it the same way (assuming that the inletted part is the missing part of course). Steps would be 1. let in the metal parts to the wood that will for the replacement portion of the stock, 2. splice that block onto the remainder of the stock - typically this is not the approach for a commercial stock but it was not uncommon on pre-WWII European military stocks, 3. Shape the replacement block down to match the existing stock and fit flush with the exposed steel parts where they meet the wood.

The model being full of weird curves might be a problem but, as long as there are no undercuts, there should not be a problem. The thing to remember is the more detailed the model the longer it will take to get a good scan of the model. Also, the files will be quite large and will probably need to be edited so it might take quite a while before you can get a usable file to carve. One other problem is the size of the detail in the item. If there is a lot of very fine detail, most probes will not be able to reproduce that part of the item because of the size of the probe tip and the small stepovers required to follow the details. You might be able to overlay some detail on the model after it is created to resolve that problem but that is just one more thing to consider.

The good thing is that once you have probed the item and edited the files you should be able to mass-produce the item without having to be right there with the duplicator moving the tracing point.

One other thing to think about would be taking the item to a business that has a digital 3D scanner that can produce a model for you. I know this is expensive but the question is, "Can the price be justified for time saved and the quality of the model they would be able to produce for you?".
Not a lot of surface detail, the curves are smooth. This is a part which I originally hand shaped, then moved to roughing out with an angle grinder, followed by a Terrco 4 spindle copy carving machine. Weird may be the wrong way to describe the curves, but since they are basically eyeballed I'm assuming they would be a bit difficult to model. One version of this I make with a stippled texture, applied with a variety of tools I've made or modified to make a random roughened texture in the wood. That is something I wasn't thinking of when I asked the question, but now that you mention it I could apply the texture as an overlay to the model if it was worth the added setup step.

I think if I do set up a CNC for this I will likely just get the part scanned. Even if the more complex part costs twice what I paid last time its still going to be worth it in time savings and I can put my attention to work holding and designing pallets to make the repeat process more efficient.

The dilemma is whether to go CNC or not in the first place. I've been making these things since around 2008 or so, and I'm down to less than an hour per part in ~50 part runs, but man is it tedious! I've found one of the bigger problems is getting consistency out of the duplicator so I have to use it for roughing out and sand to final shape. CNC would probably take longer per part, but I hope it would remove much of the tedium, it could run in the background while I do other stuff.
 

·
Registered
Joined
·
152 Posts
Thanks, it might be easier to do by hand, but the issue is that there is a "chunk" of the original stock out of the grip portion. He wants to preserve as much of the original as possible, but getting everything to mate up wll be difficult. Maybe taking a little piece to fit between the others but first cleaning up the break and getting it into a gluable surface. I am struggling with this one!!
 

·
Registered
Joined
·
44 Posts
Discussion Starter · #10 ·
I understand what you mean, generally. Obviously pictures would help with specifics, but I can give you some idea what is doable.
The amount of work involved in repairs is usually less than making a whole new stock, but for commercially made arms a new stock is still the usual solution since it better preserves value of the gun. In a case where there is sentimental value, a repair would be more the ticket.

I found a few examples on the internet. These two are repairs to Swiss K31 rifle stocks. Probably user repairs of privately owned surplus rifles. The military arsenals usually just replaced the stock with a spare if there was damage beyond minor.



This is a cosmetic repair I did, also a K31 rifle. This is the kind of work you would see the Swiss arsenals do instead of replace the stock.



Both the Swiss and the Russians also spliced stocks - but this kind of work was done when the stock was a billet of wood, before it was shaped into a stock. In theory you could do something like this on a shaped stock too, but mating two shaped pieces while keeping the alignment right gets complex. If one half the joint is a replacement piece, not yet shaped, then things get easier.

Depending on what you're dealing with, something like this could still be feasible.
 

·
Registered
Joined
·
152 Posts
I asked him to send me the photos. I no longer have the photo and he lives an hour or so away. I will post them later. Thanks for the help! The CNC might not be the best method, but our thought was to take the existing stock, determine which piece was missing, flatten and align the broken/jagged edges and try to mill something to fit. Might be easier to do by hand. I will post photos.
 

·
Registered
Joined
·
152 Posts
Okay folks! I am a newbie, but just am trying to assist this good friend. Here are some pictures of the stock and I would love to hear how you might approach this repair job.
398909

398910
398911
 

·
Registered
Joined
·
44 Posts
Discussion Starter · #13 ·
Looks like the stock bolt is bent, it should be parallel with whats left of that larger round hole in the broken section. He'll need a new one unless it can be straightened enough to remove without damaging it too badly.

That is going to be a bit hard to repair, given the nature of the damage. If he really wants to keep the old stock, I would cut off everything forward of the break and try to finger joint a new block of walnut onto the front, then re-do the inletting to fit to the action and blend the shape in from the metal back to the original stock and try to match the color after cleaning up the old part.

You could also flatten the two planes of break and try to glue on a funny shaped block and go from there. The vertical plane is obvious, then you need to go over the top to and replace some of the wood on the other side. If you can get good tight fits that can be done with two simple flat pieces rather than a single funny shaped one, but it will be obvious from the grain that it's two pieces. I see some old glue and the remnants of a hole from a pin or something on the right side of what's left, that is an old repair that needs to go too. More time required for this approach, but it is arguably stronger.

Just for good measure, it would be worth checking at Numrich gun parts corp's website for a replacement. If the gun is not especially high end you might be able to get a reproduction stock for around $100 to $200, as well as a new stock bolt. Their site is pretty easily searchable with make and model info. Sometimes they have even weird old stuff you don't expect.

We may be boring all the CNC guys here :D Feel free to PM me if you prefer. I can give you some pointers on how to do the inletting. I'm not vastly experienced with that kind of work, but I can probably get you started off on the right foot.
 

·
Registered
Joined
·
152 Posts
Thanks, will do
 

·
Registered
Joined
·
37 Posts
If you use a CNC machine to do 3D digitizing then you will end up with a point cloud file that you will need to convert to a vector file for CNC machining. You can do a point cloud file to .obj conversion then take it to a .svg file but at some point you will need to edit the file. Digitizing files are just not that accurate, so plan on lots of tinkering. I would suggest starting from scratch and modeling the stock in Fusion 360 using sweeps and profiles to get your exact curves. Then you will have that file forever. With a little editing you could even product a stock that fits better than the original. The way that break shows up in the picture is seems someone used the shotgun for a baseball bat, so check for splitting or cracks with a magnifier of some sort.
 

·
Registered
Joined
·
152 Posts
I think that the best option is to make a new stock. Thanks everyone.
 

·
Registered
Joined
·
44 Posts
Discussion Starter · #17 ·
If you use a CNC machine to do 3D digitizing then you will end up with a point cloud file that you will need to convert to a vector file for CNC machining. You can do a point cloud file to .obj conversion then take it to a .svg file but at some point you will need to edit the file. Digitizing files are just not that accurate, so plan on lots of tinkering. I would suggest starting from scratch and modeling the stock in Fusion 360 using sweeps and profiles to get your exact curves. Then you will have that file forever. With a little editing you could even product a stock that fits better than the original. The way that break shows up in the picture is seems someone used the shotgun for a baseball bat, so check for splitting or cracks with a magnifier of some sort.
I once had a fore end stock I made sitting on the bench at work (in the machine shop of a telescope I worked for) - a kid comes in, grad student who was there for an instrument install and fancied himself a machinist. He looks at the fore end and is just amazed, asking what kind of CNC I made it on. He would hardly credit that I had made it by hand. To me, CNC is like word processing or CAD plans - it saves you time making edits and not having to manually type or draw the whole thing over again. I expect the big time commitment of the modeling step is about the same or even perhaps longer than if I just made a single part by hand - the difference being that I only do that modeling work the one time and then benefit from the method by running off copies.

The point of this is that for a neophyte, yes, a CNC fitted stock might be better. For a one off I could do a better one faster by hand, and while I'm good, I don't claim to be an expert. The interest in probes is the promise of flattening the learning curve of modeling, just like the CNC itself flattens the learning curve in woodworking. But the wall I always run into is that all I would be doing is trading a curve that I enjoy for another, one that reminds me rather a lot of work (I am not a fan of hours spent at the computer - that is my day job).

In my specific case, I make a pistol grip that I sell. Since about 2009 I've made a few thousand of them and it is no longer interesting. I've been looking at CNC to relieve that tedium, but I honestly don't see that it is worth it. These things took around 4 hours when I first started, I now have them under an hour and achieve a better result. From what I understand, I could probably eliminate the worst of the hand shaping work, but production would be slower. I would trade the tedium of spindle sanding for the tedium of watching a machine run. If the CNC were a turnkey setup I would do it. But I'm faced with having to model the thing, then have to build pallets for bulk production, and then have to accept an aggregate per part time of probably 2 hours per part vs the current 45 minutes. The arithmetic just doesn't work out.
 

·
Registered
Joined
·
602 Posts
and then have to accept an aggregate per part time of probably 2 hours per part vs the current 45 minutes. The arithmetic just doesn't work out.
Assuming it's a small grip, 2 hours seems like a long time.
Also, if the machine takes 2 hours, it takes you 0 hours, so the arithmetic does actually work out. The part may take longer, but it takes no man hours.

Or, you get a machine with 3 or 4 spindles, making 3-4 parts at a time. That changes to 30-40 minutes/part at 2 hours, or even less
 

·
Registered
Joined
·
44 Posts
Discussion Starter · #19 ·
You're right that time is not so relevant if I can set it up to run autonomously. There's just too many unknowns for my comfort level with the upfront investment of time and money.

I'm basing the 2 hour guess on seeing videos of intricate carvings that take 6+ hours, but really I have no idea. Currently it takes me 12 minutes or so to rough out 4 on the duplicator, then about 20 minutes of sanding for each. Its fast, but tedious to do 50. If I could eliminate the bulk of the sanding, 30-40 minutes per part would be well worth the added cost of multiple spindles.

Multiple spindles is something I was thinking of yesterday too, that is the same advantage my manual duplicator provides, with much the same drawback - keeping everything synchronized. That would be less of a problem with CNC however, most of the adjustment slips with the manual machine come from inadvertent heavy cuts.

I should follow up with the company that did the scan of the rubber part I had made. $500 for a .stp file is pretty reasonable when one is selling stuff and has no idea how to model a part. Do the usual CAM programs used produce Gcode from .stp's?
 

·
Registered
Joined
·
37 Posts
The .stp file can be imported into Fusion 360 modified and then saved out as a svg, obj or other CAM type of file. When I use my CNC machine to make parts for our local woodworkers club I usually stay in my shop in case something goes wrong. Most problems are in the setup and 1st cut after that I just let the machine run. If you have several different bits used in your project then you will need to do the changes. I would strongly suggest learning Fusion 360 your time will be well spent even though you may decide not to invest in the CNC machine. YouTube has many videos on Fusion 360 training and there are plenty of books that help also. Part of the fun of modeling with Fusion 360 is you can do what-if's while creating the drawing and check for clearances. If your into production of large quantities of product then I would look at the project size and maybe invest in 2 or 3 smaller CNC machines instead of one big one.
 
1 - 20 of 24 Posts
Top