Has anyone thought to try pressing 13/16 aluminum round into their emt rails? I think if you used a paddle wheel sander on a a long bit u could take out enough material for it to fit. I’m thinking of trying this in a few weeks. I figure I can smear the rod with some epoxy and use my drill press to press it in.
Aluminum Is much less rigid than steel. I doubt it would have much effect. There are a handful of tube stuffing threads here, but really most of the work is done by the skin of a tube.
6061 aluminum is pretty stiff, comparable to mild steel if not a little stiffer. That’s all emt is anyways is mild steel. It’s designed to be bent.
Part of what happens when emt flexs is it gets elongated at the bend point so it is not a true circle anymore. Having aluminum inside will keep it from doing that.
I suppose it wouldn’t be too hard to test it. I could place a 3’ peice in a clamp and hang weight on it, then measure deflection. Once with alu filled and once plain emt. I would try to outright bend it, but measuring deflection seems less destructive. 
A better bet is just using the 1" SS (I just got mine for $42), if you are looking for more than EMT (~$12) offers. That is running on the assumption the rails are the weak point, which totally depends on the size of your build.
I’m aiming for 2 x 4. What have people found to be the weak points on these machines? I’m working on an alternate build and it would be great to try and solve any issues in advance.
Your right, it would be easier to have used the stainless tube, but I already had emt here, and started printing before I started thinking.
I’m having troubles navigating to specific useful information in the forums. Running a search returns any post with the key words. It could help to be able to search titles only.
I spend all my time on this. I have done everything I think I can, the second I think I have an improvement I release it.
The weak point depends on your use, usually the complaint is speed but I think that is the complaint for every cnc in existence for the cost they seem like they should go faster. For such a giant build why not make the LowRider, it was designed for this very reason, to be larger than the MPCNC?
It is not worth the effort to compete with google, no search I can implement will ever be that good. You can use it to just search the site.
I think I should point you here. There is a common misconception that a 3/4 tube is stronger than a 3/4 bar stock.
There are some formulas in there that can help calculate filling with alu may add 2.8 times the tensil strength.
Additionally I’d like to question your excitement of adding 1/2 tubing in the free space of many components.
I have held both 13/16 alu and 1/2 tubing and the 13/16 is by far less flexible.
Adding 1/2 hardend square stock may prove useful.
There is no doubt in my mind bar stock is stronger. We have all done lots of math on this. I would love for you to try it out. I am not trying to discourage you in any way.
One thing I am sure you can easily agree on is more mass equals less acceleration for the same system of rigidity. So the heavier you go the more rigid it needs to be…see where this is going? I think you will find this is a properly balanced system as is. 1 part does not equal the whole. Think of it this way, will a 10HP @ 50lbs spindle help or hinder, what about a .0001HP @ 1 oz? debating small amounts is harder to see but if you go to the extreme it gets a littel easier to see, a small change in rigidity, and a small change in mass, may or may not equal out. What about just using a single flute endmill to cut the load in half, or a four flute quadrupling the load… 1/16" vs 1/4"? There are easier ways to chase cutting speed, because in the end that is all you are trying to do.
When you build it I am sure you will have all sorts of ideas, that is kinda the point, to have a CNC machine that you can mess with. Common parts and sizes, inexpensive and easily to manipulate. I have done my far share of math. Try it out throw some indicators on it and let us know how it changes. Just a bit rough to try to improve something you haven’t tried. Looks can be deceiving, you could very easily be trying to change a part that is not the weakest link.
You are going to fit right in here, hurry up and build it so you can join the discussion!
2 Likes
Thanks. I am waiting on bearings. After thinking about it for awhile I figure 3’ 2" is what I would need. I’ve seen rail supports on thingiverse, I’m hoping will give it enough strength to cut through hdpe, and soft pine.
As for the modified version of the machine I’m also working on building, it’s a gantry style machine. The bed will track along the z axis, I’ve already completed the model for these parts. Y axis will move left to right along fixed rails ( I’m about 80% done with the carriage and still need to modify the motor mount to act as the feet at both ends of the rails. For the z axis I’m going to use 12mm rails and brass bushings. I wasn’t satisfied with the amount of stand of between the y and z axis.
So essentially it will resemble the other mill or carvey. Only I’ll be using your solution for cheap linear rails. In this application the weight of x and y rails will have no effect.
I have not 100% decided on belts or screws. I’m thinking dual screws on the x axis. Advise in this department would be appreciated.
This is the x axis test print. It’s completely redrawn for the bearings I already have. They are nmb l-1680hh. Essentially there will be four carrying the spoilboard. I’ve inset the nuts to get and exact fit for the bolts.
Y axis and z bearing housing is intended to be one piece, and rails will be bolted through a piece of 5/8 square stock, 1/8 steel , and 1" plywood.
The intended cutting area is 1’ x 1’ x 5 1/2. I will post all designs once it is completed.
Also the spindle I intend on using. I have some pulleys coming to get ~ 6000, 12000, and 18000 rpms depending on which set is used. Also im hoping for software speed control. Hopefully I can get it to work closed loop. I have a hall sensor coming, not sure how well they work at high rpms.
Anyways thanks for your hard work. I’m hoping to help build upon what’s been done and add to the home brew arsenal. Milling aluminum to good tolerances is this machines goal.
Can’t wait to see it.
With a moving bed don’t forget to account for material hold down clamps/screws. That will also drastically increase it’s footprint as typically we have material hanging off the edge of the machine, one of the advantages of not enclosing any of the sides of the MPCNC.
Belts, 1000% belts. A ball screw is the only other way to go but maintenance and cost easily make that a poor choice.
It is best not to sink the nuts. They vary in actual size all over the world. With each revision I make I have been removing that “feature”. My first parts were super easy to assemble…for me, but outside the US it was a nightmare. Best to just start without doing that so everyone can use it. Any screw or bolt needs a metric equvailent, so you are pretty much stuck with 5/16, 8mm and 6/32", 3.5mm (or 4mm if you make it loose).
Well I can go on for days about all this but there is a reason for every design choice I have made (or been forced into). My only advice is don’t don’t make component decisions for yourself, make it available for everyone. So many good builds are difficult because people use random parts from the scrap bin.
1 Like
I see you have put a lot of thought into this design. It’s the main reason I choose to emulate aspects into my design. At the moment I’m not worried about the metric parameters, I just need a working example.
It takes about 3 hours for the x axis parts and currently 5 for y and z. I have limited time to actually draw but I’ve been thinking about it for days if not weeks.
I don’t plan on deviating from your design much on my mpcnc, being said I will be using double y rails to comp for 3’ length.
For the enclosed design I spent a bit of time testing deflection of mounted rails. With the frequency of taps and mounting platform, theoretically it could make a light pass on steel. I plan on attempting it to identify weak points.
I will make a new post in builds. I think I should provide dxf for all plywood as well, but I will be doing a good portion by hand.
Ryan, what do you use for design program. I’m using sketchup. It’s kinda wonky but really easy. I used to use blender but I spent too much time writing my own modules that I just got burnt out on it.
Thanks for advise on hold down. The plan is to minimize available hold down on x axis and utilize extra space on y.
Ok, so there are two inexpensive ways one could stiffen the frame without adding much weight.
The first is to use a truss method. In order to do this you would print a spacer to hold the all thread to the inside bottom of the pipe in the center. You need two more spacers at each end. Use two washers larger than the inside of the pipe and some nylock nuts to add pressure. This will add rigidity in one direction.
Another method would be to use 1/8 x 3/4 inch carbon bar stock. Grind it down to 13/16 and insert it inside the rail with a little bit of epoxy. 3d printed spacers would come in handy here as well.
Bed rail is hardend steel, and if you buy a harbor freight grinder guide table, you can make a pretty accurate strait cut for cheap.
Both methods can double or more the tensile strength of tubing without adding too much weight.
On your machine when you wiggle it from the tip of the endmill what moves?
You should try them out, that is a bold claim. A member under compression, vs tension, at what load does this overcome the steel and start to rely on the CF (total deflection)? Again, you are assuming the rail is the weak link, why?
Thought exercise. Infinitely rigid frame, now what? Infinitely rigid machine, how much speed/accuracy do you gain before you have to spend more on a spindle or larger more expensive endmills?
We are at a point with the 1"OD stainless that under normal conditions average size it is not the issue, not even close.
I’m using a larger foot print on my biuld and using a custom lightweight spindle.
For my case use I can afford some additional weight in the rails if it provides more rigidity across a slightly longer span.
FTFY
Bed rails are made of whatever scrap the steel mill had at that moment. I wouldn’t call it hardened, but it can and usually does have hard spots in it. I had a single center span support on my 5 foot square mpcnc, worked great. The issue was the moving gantry pipes sagged a little under load. I didn’t bother with nuts inside the pipe, just a nylock cranked up to the tube with a little of the allthread going inside the pipe to keep it from wiggling sideways. The tube never left the nut, both when cutting a few thousand holes in some mdf with a V-bit, or carving aluminum.