Commit 0b3f0dad authored by Ruben Castro's avatar Ruben Castro
Browse files

added notes

parent 44e4daf0
This diff is collapsed.
# CAT 0030
## Goals:
Differing from previous machine:
- 24"x18"x3" travel
- Tallness parameters
- Routing plastic
- Motor mounting with back support to prevent bending when belt tightening.
- Delrin Everywhere possible. Maybe Acrylic rails?
- Mounting shit to bottom should be easier.
## After building the first machine pretty carelessly. Let's use math.
#### Max possible loading.
I want to be able to machine 1/2 Delrin with this machine with a few passes.
I don't care as much about too much deflection in the translational manners. Rotationally though, that matters more I think. That's why I really want to use aluminum extrusion but alas...
We can account for translational deflection using software I feel like. This would be a cool project in in of its own. That would actually be super duper awesome software. It makes Delrin that much more plausible.
#### example
looking for examples for routers that cut plastic and wood, there is two I like mainly:
- X-Carved
- Shapeoko 3 XXL
Reach: ShopBot Max
##### Shapeoko 3 XXL
- Seems to use same GT2 belt. Pretty much same belt routing design as Jake's
- Uses V groove bearings on aluminum extrusion. Attached through a 1/8"-lookin steel plate.
- Uses NEMA 23 motors.
- The motors are on the machine
##### Vibrations, Beefyness and Inertia
Cast Iron is used in a lot of machines apparently because its better for dampening vibrations whilst routing parts. I wonder how much of an effect the vibrations will have on this machine. I can account for static loads and deformations through software but if we get to a resonant point, will everything just get rekt? Something to keep in mind for the future.
As seen on Shapeoko, to increase inertia, motors are on the carriage as opposed to static on the bed. I'm thinking maybe if the machine will be a router, have ability to mount motors onto the actual carriage? The increase in inertia could actually be a good thing for a CNC router type machine.
Time to do some matttths.
## Prototyping
### V-rails
V rails are awesome but the beam joinery really starts to take up space on the cross webbings. I'm trying to determine okay ratio of lip of PLA beam joinery vs strength of Delrin for the end grasps.
PLA Tensile strength : ~7080psi, Delrin Tensile Strength: ~10000 psi
I like this idea a lot. It also allows the bed to have diagonal rails and it would not take up too much more space. This is dope idea.
P.S. There is like no one in the lab today. It's just me. Me and so much lab equipment. So much power.
For bearing tightening.
### V-Groove bearings
To mimic a little bit of the Shapeoko, V-groove bearings can save complexity and the number of bearings. I want to try 3D-printing a V-groove bearing that sits on two bearings.
Maybe buy these?
I don't like this that much because in that case, some of the loading would be axial as opposed to radial. When it comes to cheap bearings... radian loadings are much better? Complexity is also not too much different after design as well. Maybe for a future low-load machine it might be worth a try.
## Maths on weak points
From just eyeing it, you can identify a few weakpoints.
1. Carriage Plates holding onto the x axis carriage and the bearings. This is a 1/4" plate that on regular machines is regularly made of stainless steel.
2. X axis torsion. This is a long beam and torsion might be of quite a bit of concern with x torque.
3. Z axis torsion. This one goes along with X axis torsion.
#### Carriage Plates
one of the weak points is the side plates. According to my calculations by ignoring shear stress ( calculated to be 2 orders of magnitude less than bending stress), the maximum stress on a 1/4" plate used to attach carriage to bed would be around 1.95x10^5 Pa. On a 1/8" plate it would be around 4x that at ~8x10^5 Pa. This means our 1/4" Delrin plate received around 1/4 of the stresses of the steel plate used by the Shapeoko router. Stainless steel has a tensile strength of around 505 MPa, whereas Delrin is at ~ 69Mpa. This means our Delrin is (69/505 * 4) = 54% as good as the 1/8" stainless steel plate. This is doable considering the steel plate is meant for machining aluminum. This is good news.
#### X axis on Torsion & Z axis
After simplfying models, there is an estimated 0.63 cm deflection at the tool tip coming from the Z axis being a flat sheet simply supported at the top, and a 0.21 cm deflection at the tool tip due to torsion of the whole X axis. The main fix is to strenghten the Z axis.
Shapeoko uses a two sheet design connected with standoffs for the Z axis. Could do this and when routing, maybe the tool is what goes underneath? tooling is steel, so it would be much better at resisting those bending moments than the delrin.
Also... Spring on Z axis?
#### Y axis Bearing Rails. I'm not sure how I feel about these.
Gettin better.
#### New bearing design.
With the 45 degree bearings, tightening was easy because contact was only at a point. There is no need to maintain alignment. With the full contact bearing, there is now extra bolts needed to tighten it up. This adds up to more space which is fine and attainable for the y axis carriage attachments. But I wonder if there is a neat solution for the x axis to take up less space. Maybe one is enough and let flexture and the pressure from the material makeup the difference?
##### Mounting shiz to bottom
Im looking into solutions to more easily mount on to the beds. Large stuff should be mounted using the .
JK. M3 Heat inserts totally go in 1/8" no problem easy boiz.
##### Bed on top of Bed Problem.
##### Side Walls
Problem here is that the side wall can no longer be one continuous piece unless you hyper extend the rail which is just no fun. So I'm trying to come up with best singles design and I'm having trouble deciding if a single mounting attachment is enough.
Models are run as follows:
All runs have x axis constraint on face.
1. Y axis constraint on all tabs, and Z axis constraint on the 3d print. 40 N y axis load on the top tabs.
2. Y axis constraint on all tabs, and Z axis constraint on 3d print **and the bottom face**. 40N y axis load on all the top tabs.
3. Y axis constraint on all tabs **and 3d printed connector**. Z axis constraint on 3d print **and the bottom face**. 40 N y axis load on all the top tabs.
I want to use this as a preliminary test to come up with a basic design to choose with the understanding that this test is flawed and will be redone later with more design elements being modeled in.
Model types are the following:
A- 1 tab on top middle, 2 tabs on bottom sides and 3d connector on bottom.
B - 1 tab on top middle, 1 tab on bottom middle and 2 3d connectors on bottom sides.
C - 2 tabs on top and bottom sides, 1 3d print connector on bottom.
D - 1 tab on top middle, 2 tabs on bottom sides, 3d print connector on bottom, and two short tabs on sides. ( for this one, all three runs contain Z axis constraint on side tabs)
These models don't take in connecting each unit to each other. That is something I can test later if the results come back severe from analysis.
Highest MPa.
Test | A | B | C | D(2mm/4mm relief or just 2mm)
1 | 93.81 | 8.641 | 20.2| 6.675/8.701
2 | 8.796 | 8.641 | 6.229|8.699
3 | 8.795 | 8.640 | 6.23|8.699
Highest Deformation (mm)
Test | A | B | C | D
1 | 4.374 | 0.001 | 0.01 |0.001/0.001
2 | 0.072 | 0.001 | 0.001| 0.001/0.001
3 | 0.058 | 0.001 | 0.001|0.001 (6.6x10^-4 actually lowest of all)
This pic is of Simulation D,3
B,1 has this super high stress in the 3d print connector. This makes sense because it is the only thing stopping the rotation and it is inline vertically with the load. This is why B,2 shows much easier results on the part, the z axis support from face helps a lot.
Highest stresses are on top tab with tests 2&3, while most stresses are at the bottom 3d print connector for Test 1.
A main goal of this machine is to be easy to remake. As such, I want to limit the top face to only 1 tab, as those tabs need to have a 3d printed fitting for the twisted rail. To minimize parts and complexity then, we would like to have designs A or D. Design A does not have enough stiffness to maintain a good machine. As such, design D offers the easiest, and still formidable option, being nearly equal to options B and C. Anything more would be overkill.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment