Round Stainless Tube + Plywood + 3D Print Router - Design Critique

I posted a while back and got some great feedback (Thanks Peteeng), and finally decided that the PrintNC would be the best design for me. However, after much investigation, decisions on required work area, sourcing steel tube, etc., I came to the conclusion that I don't have enough free space available for a router/mill capable of my desired 24" x 24" work area. I would prefer not to scale that design down too much as the cost does not decrease significantly, and an upgrade (scaling up) later would cost considerably more (new linear rails and ball screws, as well as additional lengths of steel tube) on top of the price of building the first machine.

The MPCNC is recommended by a lot of 3D printer enthusiasts, but I often hear that this machine is quickly outgrown and is limited by the rigidity of the plastic parts acting as structural components. I found the Solsylva design, which caught my interest since it used bearings on pipe, while using plywood for rigidity. I took aspects of this design, coupled with the MPCNC and a bit of Onefinity inspiration (I only found out about this machine after I had drafted up the design).

What am I missing with this design (picture attached)? Quick summary:

Expected material cutting capabilities: plywood, hardwood, aluminum
Red parts would be 3d printed, with plans to mill replacement pieces out of aluminum.
The wood-coloured parts are plywood (surprise!), but could also be cut out of aluminum later. The holes seen in the upright boards, where the rods appear to be held in place would actually be clearance holes, with 3D printed clamps (bolted to the plywood) to locate and attach the tubes to the plywood, as well as allow for some adjustment.
Bolts and clamps would be used for fastening as opposed to wood screws (I have plans, but I want to keep this relatively brief)
The linear 'rods' are stainless tube, with plans to use 1", 0.125" wall thickness 304 stainless
The 'bearings' are three 608 bearings arranged in equidistant triangular pattern around the tube, with six bearings total contacting each tube (three at the front, three at the back).
The planned work area is ~16"x16", with a Z of 2-3"
Spindle: 800 watt router, or 800+ watt water-cooled spindle w/vfd
Drive system: TBD, but belts or ballscrews are the current top options, dependant on the rigidity of the frame

You need to add a lot more detail,and you probably know that.Is there a reason for using printed components,other than owning a printer?You need to model a spindle or router in there as the travel will be reduced by the diameter of the body relative to the distance between the upright ends.Will the machine be moved by belts or screws?

Thanks for the feedback. I will try to address each of your points.

1) Yes, this design is still in the early stages, but I was looking for feedback on the mechanical rigidity of the basic structure. This concept may have been done already (I wasn't able to find a suitably similar post), but users may have seen a similar concept succeed or fail for various reasons. The picture was thrown together in a few minutes in Fusion, so it is definitely missing a lot of details.
2) The use of a 3D printer will enable me to create 'complex' geometries that may be better suited to machined parts since I do not have access to a mill or router (yet)
3) I didn't model the spindle because the structural rigidity of the frame should not be affected by the spindle. I will definitely take the diameter of the body into account when finalizing the frame dimensions - Thanks for the pointer.
4) I addressed this in my final point on my original post, but it probably got lost in the mass of text. Belts are the cheaper option, but I would prefer to use screws for the obvious reasons of lack of belt stretch, mechanical advantage, etc.

Hi 147 - Whats Missing? 1) rigidity 2) SS tube is not intended for bearing rails so its tolerance and surface finish are poor 3) bushes suffer from a thing called stick slip. Getting four bushes and two rails to co-operate is a difficult thing to do. Bushes require clearance to work and this clearance creates chatter and stick slip. Its the same as an F clamp. You can adjust the F clamp and once cocked it stays. This is what will happen to your bearing arrangement 4) The bushes length is less then the dia of the tube. Further issues with stick slip... 5) the SS tubes have to be perfectly parallel to work. Due to the tubes not being straight, round or polished stick slip will get you again. Rolling elements are used for very good reasons in machines....

In summary the cost of cheap linear bearings is worth it... Developing your own bearings is poor use of your time and $$$ even though you think you don't have $$$ you will spend many many $$$ in failures going down this path.... Peter

https://en.wikipedia.org/wiki/Stick-slip_phenomenon

HI 147 - just reread your post and you are going to use bearings around the tubes. This is good but by the time you add up the bearings cost I think you will be at the same cost as a cheap linear bearing... I'll leave above for reference material.... The twin rail design is redundant on the Y axis . Better to use a much bigger single tube then two small ones... The gantry needs two but due to the tubes poor tolerance (straightness and roundness) you will have tight spots and loose spots. Your best bet here is to use hydraulic rod material. This ground to a tolerance and straightness. Peter

Thanks for the reply Pete.

To summarize my points below: This is meant to be a redesign based on a pre-existing BOM/design, which was originally built to a price (if I understand the scope of the original design correctly). There are better value designs and better value linear motion systems available, but I was looking to improve on an existing design by focusing on some of the weak points.

Linear bearings or better yet, linear rails are without a doubt a much better alternative, and for the machine that I am planning for myself, linear rails will be used. However, the above design is meant to be an improvement on an existing design, by re-using the original design's linear motion system. It's a compromise, but the purpose is to re-use as much of the original design's components as possible so that this design will be a 're-build' of sorts for current owners of the original design, rather than a new build altogether.

I'm familiar with "stic-tion" in bushings and I try to avoid them for exactly this reason. I have a mechanical background so this was a subject that was studied to some extent. However, using linear bearings would require the use of hardened (or reasonably hard) round tube or rod, which can get expensive for long lengths. Alternatively, the tube/rod can be used as a consumable, but this has it's own costs and setup issues. I've never thought of using hydraulic rod material before, so I'll definitely have to consider that in the future. My current go-to for linear bearings (light load applications) is O1 drill rod, but I might be able to harden it for slightly higher load applications.

The use of four Y-axis tubes (two per side) was something that I wasn't sure about. I've seen machines the just use one Y-axis rail per side (two total), but some complaints about these machines is the lack of rigidity. Some of the more rigid machines use two parallel rods per side. My thought process was that the two parallel rods will help resist moment/torque about the Y-axis, but I don't have a direct real-world comparison to validate this theory.

I checked the ASTM A554 spec for welded Stainless round tube, and it specifies a maximum of 0.030" straighness deviation per three feet (if I remember correctly), which is reasonable for this application (a very low cost router). Again, this is not the best option available, but it is mostly just to re-use the existing BOM.

I agree with peteeng,nothing new there.Have you looked at using SBR type rails ? https://www.redcnc.com/p-supported-linear-rail-8 .Not enormously expensive and a huge step up from tubes.Then you could add the triangulation that your end rail supports have to the uprights at the ends of the gantry,which with a plywood box section beam across the gantry would get you a much better frame.It might even save a chunk of time as 3D printing isn't exactly rapid prototyping if using a home printer.