This is my attempt to explore some ideas with feedback (hopefully) from people more experienced than me with the lowrider. But this could theoretically be applied to the MPCNC as well.
The idea: Engineering 101 says everything is a spring. You just have to quantify the degree and direction based on the forces at play. So as I’m playing with my lowrider v3, I got to thinking about how to improve the rigidity of the conduit to reduce torsion forces from the cut itself and gravity. I started thinking about how concrete is post-tensioned with cables stretched under load to apply compressive force to a slab and prevent deformation under load.
The question: Has anyone tried tensioning a cable (more likely a piece of all-thread) with spacers at roughly the same intervals as the printed braces to take up the gap in the conduit and the cable, to specifically prevent, or at least discourage deflection from gravity/the mass of the tool? I dont think it would eliminate it, but i think it could help. It also wouldnt do anything about torsion on the gantry, but it wouldnt add much weight.
Just thinking out loud, on the internet (a dangerous practice, to be sure). What say you gurus of the interwebs?
To be explicit, I mean a washer on each end with all-thread through the center of each piece of conduit, with printed spacers and jam nuts spaced evenly across the conduit, tightened til Lord Helmet goes to plaid, then the ends trimmed flush with the nuts on the end.
There have been a lot of experiments about filling the tubes with various things to try and get them more rigid. Searching the forums should get you the full story on them. From my recollection, there was always cost and complexity added, and very rarely a measureable increase in rigidity commensurate with that cost and complexity.
My TLDR summary is that the juice was never worth the squeeze.
I am honestly not exactly sure how that would play out. My gut says the tubing is not strong enough to get enough tension in the all thread to make a measurable difference. The same amount of added mass to a larger diameter or thicker walls would probably far outperform the tensioned rod.
I would love for you to give it a try though. I have had a dial indicator all over my LR4 beam lately and it never fails to surprise me about how wrong I can be.
Right, but I’m not looking to add rigidity by stuffing the tube with something. Im talking about applying force to a secondary member in the system that then supports the tubing from the inside.
Admittedly I have a few irons in the fire at the moment, but I will likely try it just for the curiosity of it. Worst thing that can happen is it doesn’t work. Best case scenario,we get more rigidity with minimal added mass and the simplest of construction.
So I absolutely could be very wrong here, we have entertained this idea a lot, and I know someone has actually tried it, but years ago, so please do not take this as discouragement. I would love to see another test. I will offer my perspective, and again, it has been a while since I studied this but here is my understanding.
The issue here is a central member adds almost nothing, even if it is stretched very tight. .02mm on the outer skin ads magnitudes more. Thinking in terms of a solid beam, top in tension bottom in compression center is zero. Then you add on the next part.
I think most people associate this with post tensioned concrete and that is doing something very different. That is compressing it and making it more rigid. With this you are trying to use the tube itself to make a central rod tight and extra stiff, then transfer that stiffness back out. So you would be adding a very high load to the outerskin and possibly adding to the buckling. The ends will need to be extremely parallel, or you will add in a bow.
Now if you could tension that outside the system then transfer the rigidity out it would 100% work but in this case I think it is sort of like trying to add a fan to a sailboat. You can’t push from the item being pushed.
So that is my assumptions. I would 1000% love to be wrong. I am trying to make that beam a bit more rigid, and have been analyzing so many aspects of it…but not this one.
Well to add to the cons side of this idea, the shorter the gantry is the less this effect would matter if it works. So this is really only relevant to the longer gantry machines. But as I do my testing, I’ll test a few lengths. If you want to point me in the direction of the most common lengths, I’ll include them.
Since I’m only looking for deflection, I plan to isolate the tube to remove other factors, then point load the center with a mass equal to the mass of the carriage with a tool plus 10% (because over-engineered is better ). I’ll probably load the carriage models in the slicer to get a total mass printed solid, then just add the weight of my router. If I test the worst case, it can only get better from there, right?
The weight of the tool and core should be trivial.
The largest load is from cutting and it is a torque. direct x, y and z loads are handled well by the strut plates from what I can tell.
if you are going to test it you might as well just work on the full width beam that is most common to see the most difference. smaller will just be diminishing returns
To be honest this all kinda stemmed from the pattern I got trying to surface my spoil board. I got a 40thou difference in z from X0 to Xmax. Which i know can be fixed by adjusting the z ends tops to make them level. But l also had a 13thou concavity in the center. The 10-11thou lean in the cut along the Y axis was probably me being too aggressive with the cut (60% step over with a 3/4 bit and a 0.050 DOC.
Yeah the software leveling would have handled the 1mm of difference pretty easily. But now that you machined your surface it is done.
3/4" (19mm) endmill is extremly large. We typically use a 1/2" (12mm) or less.
60% stepover is also large and par that with a 3/4" endmill and a 1.2mm DOC. You were taking of what I would consider and EXTREME amount of material per revolution. A surfacing pass is a precision operation not a roughing pass. I do my finishing pass in 0.2mm increments. 40% stepover max.
Do your surfacing again with a reasonable endmill and a 0.2mm DOC and see how flat your surface becomes.
The same hold true for any cuts. I can put a 1/4" single flute on and do a roughing pass and it will pull the machine down. That is why you always need a finishing pass on all surfaces you want to measure.
Right, I get the parts that were my fault. I set my expectations too high coming from what I would consider a very light cut with my mill. If it only costs me some MDF to find the limits, then that’s ok.
Now that I know the errors on my part and some of the limits of the machine, I’m trying to find a way to increase those limits or at least mitigate them.
I think you should start with a standard cut and work from there. Obviously I think we are already extremely balanced, but that is for the typical use case. I don’t know what your use case so I am not sure how optimized we are for it. But trying to optimize for a once a season extreme surfacing run would just be a fun experiment, not what I would want to spend much of my time on.
But a 1/16-1/4" single flute up cut in foam, wood, plastic should be near the end mill ideal range for MRR. This is what the machine is designed for. Surfacing raw slabs (larger diamter endmill) will be slower than an Avid or something similar, but the bang for buck we would still be a clear winner (you could be running several LR4’s for that price, still with one operator, and easily have a higher MRR at the expense of floor space).
My ideal use case for the machine is largely centered around graphite molds for aluminum casting.
In my shop I almost never make the same thing twice. I just fix whatever gets brought to me. Being able to easily and cheaply remanufacture a casting from a scan would help me and the people I work for tremendously.
What is the largest one you will be making? Roughly?
The smaller the machine the more rigid it is, dramatically.
So I am assuming you will be using small tapered ball end mills to do the finishing passes? I do not for see you having any issues at all. The cutting loads from a 3/4" at 1mm DOC, and a 1/16" ball end 5% step over are opposite ends of the spectrum.
Excellent question. I wish I knew. Every time I think I have big enough, or even enough machines to do the jobs I get asked to do, I find out I was wrong. So I tend to lean towards as big as possible with the work envelope.
Yes, primarily ball end mills for finishing, but it depends on the part/feature. All the finish machining will be done on the big cast iron machines.
). I’ll probably load the carriage models in the slicer to get a total mass printed solid, then just add the weight of my router. If I test the worst case, it can only get better from there, right?