Ideas for the next Wally.

Re: Ideas for the next Wally.

Postby Nicholas Seward » Fri Apr 10, 2015 9:38 pm

@realthor: Your idea would work. I would make the pulleys on the figure-8 bigger. The easiest way to drive bed/top would be a screw driven from the top or bottom plate. In addition to what see3D mentioned, the reason I suggested the other design was to minimize the number of steppers that aren't static. For every stepper that moves there are wires that will be flexing. With careful planning and prep this shouldn't be too much of an issue.

Note: My last design could be wall mounted. You could also get a Z travel that is almost double your Y before your arms start sticking past the bed. However, you want to keep the arms shorter so they don't have to be as beefy to combat the flex.

Wally Bed2.png
Wally Bed2.png (40.85 KiB) Viewed 13532 times
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Ideas for the next Wally.

Postby Nicholas Seward » Fri Apr 10, 2015 9:50 pm

@realthor: What see3d is talking about is... Any time you have the effector accelerate toward or away from the wall and the Z arms are not horizontal then the arms will get forced up or down slightly. This is even worse if you tie the top and bottom together which will double the effect.

mass_effector*max_acceleration_effector=reaction_force

To engineer this you will model your arm system and apply that reaction_force. You would increase the rigidity of the components until you get an acceptably small amount of flex. I rarely design right at this edge and just go for overkill in the interest of time and standard components.

Bottomline: the more mass that is static ... the less you have to worry about dynamic consequences.
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Ideas for the next Wally.

Postby see3d » Fri Apr 10, 2015 10:10 pm

At this point, without deviating too much from the original Wally design, I would carefully consider mounting the elbow stepper directly on top of the shoulder pulley. I believe this would considerably simplify the calculations. The fact that the elbow stepper is rotating with the upper arm should not be too bad for inertia and the static load is not in a position to create a bending force, as it would if it were out on the arm a ways. The wires would have a small range of linear motion also.
see3d
 
Posts: 67
Joined: Tue Dec 03, 2013 1:24 am

Re: Ideas for the next Wally.

Postby Nicholas Seward » Fri Apr 10, 2015 11:15 pm

@see3d: I agree.

WALLY ARMS.png
WALLY ARMS.png (59.45 KiB) Viewed 13525 times


To further muddy the waters, here is a completely new approach to the arms. (This is stolen from how I am doing my 1-arm SCARA.) The shoulder pulleys are free wheeling. The elbow pulleys are attached to the forearms. The motor rotation will now be directly controlling the forearm angles. This takes the math down to one trig function and one square root among other simpler operations. The forward kinematics is also now possible.

However, at this point you might as well have a 1-arm SCARA. Have one stepper drive the shoulder with an attached pulley and have the other stepper go to a free wheeling pulley as shown now. Strangely enough, this makes the math more complex (Barely. Two trig function instead of 1.) but would simplify the mechanics (2 arms instead of 4.) and would increase the build area.
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Ideas for the next Wally.

Postby Nicholas Seward » Fri Apr 10, 2015 11:30 pm

Since I am in the crazy idea share mode... I have thought about making a bot that instead of being mounted to a wall would lay on the ground. The Z could now be made linear by just putting a few rollers on box. People will freak out because you are printing sideways but it is better for many reasons. This will fix the two biggest issues with a bot like this: arm droop, bed droop.
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Ideas for the next Wally.

Postby see3d » Sat Apr 11, 2015 12:40 am

Hey boss, I like the crazy idea except for two things:

1. Will the tall heavy print stay attached to the bed?
2. Many things become overhangs that used to print fine. Think hollow vase.

This is kind of like a moving Y with a right angle bed attached to it and then a CoreXZ for the XY, only with a single arm SCARA.

When I mentioned the motor sitting on top of the shoulder pulley, I was actually describing driving it as a one arm SCARA (thinking of another version of Wally that we had described a while back). The other arms were just for stability (in theory), but perhaps they are not needed in practice. I just know that a card stock paper model held the arms rigid when done Wally style, and it was flimsy with a single arm.
see3d
 
Posts: 67
Joined: Tue Dec 03, 2013 1:24 am

Re: Ideas for the next Wally.

Postby Nicholas Seward » Sat Apr 11, 2015 3:17 am

@see3d:

1) If you have a heated bed that is properly prepared this shouldn't be an issue. However, I want to get away from heated beds. I have seen people do cold glass beds coated in tons of hairspray for PLA and it can stick so good that you have to take the bed to the freezer. I will have to try.
2) Turn your print on its side and print a hollow vase. I think you will be surprised that nothing changes. gravity_forces<<plastic_forces.
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Ideas for the next Wally.

Postby see3d » Sat Apr 11, 2015 4:20 am

Unfortunately, I can't turn my printer on its side. It is balanced against gravity. Z uses gravity to stay down, X and Y assume no gravity bias. I have a really big heavy printer. I attached a PEI sheet to my aluminum heated bed. When it is warm (60C) PLA sticks really well. When it cools down it will release it. It does not wear out.

The Tiko KS printer claims that they can print PLA and it sticks to their flexible bed well. They give it a twist to remove the print. No heated bed option. They won't say what it is made out of. They are going for the cheapest printer $179 can buy.

There are some others that are selling plastic bed material that is supposed to work without heat for PLA. Gecko Tec KS is now shipping their thin steel plates with a coating for PLA. This is a "hot" area for development right now.

If we had good FW, I think the single arm + rotary bed design is the sweetest thing I have come up with. I even gathered up a few parts for it in case I decide to build it later on when I have more time. Tiko KS is going to build their own slicer that generates code for the delta motors directly instead of G-code. That way they can cheap out on the FW processor power.
see3d
 
Posts: 67
Joined: Tue Dec 03, 2013 1:24 am

Re: Ideas for the next Wally.

Postby realthor » Sat Apr 11, 2015 1:48 pm

1)
Nicholas Seward wrote:@realthor: What see3d is talking about is... Any time you have the effector accelerate toward or away from the wall and the Z arms are not horizontal then the arms will get forced up or down slightly.


I am thinking that because of the symmetric design, if the top platform and the bottom platforms have similar weights, the effort on the motors will be minimum, as they will balance each other (that is, if the top platform is heavier it will tend to press down, lifting the bottom platform). Considering this, having the weight of the SCARA arm, the extruder, the motors /etc on the top platform would actually render the inertia from fast arm movements less perceptible.

2)
This is even worse if you tie the top and bottom together which will double the effect.


I was actually wary about the z-axis resolution due to the double amount of move per motor step. This would mean half the resolution for the same settings. I would need to consider a reduction drive gears and a high number of steps per revolution motors, maybe even microstepping in the motor drivers. I am quite lost in this space, please pitch in with any comment.

3)
To engineer this you will model your arm system and apply that reaction_force. You would increase the rigidity of the components until you get an acceptably small amount of flex. I rarely design right at this edge and just go for overkill in the interest of time and standard components.


Placing the weight on the platforms would decrease the inertia and maybe other movement defects but indeed I would need strong arms. In a rep-rap fashion, this would be achieved with a strong filament (carbon-fibre infused) and a truss structure, or, for now, a milled aluminum or even wooden structure.

Thanks.
reprap Lander concept on Concept Forge
reprap Lander concept on RepRap Forums
realthor
 
Posts: 83
Joined: Tue Mar 31, 2015 7:45 am

Re: Ideas for the next Wally.

Postby Nicholas Seward » Sat Apr 11, 2015 9:08 pm

@realthor:

1) The weight of the steppers will help you but the effect will remain. Testing or calculations will have to be done to see if the effect is minimized enough.

2) A simple screw should do it. Figure out what your target resolution is and make sure you can hit it before you get into microstepping. Microstepping helps but 1/32 microstepping doesn't make your resolution 32 times as good. (The position can be +/- 1/2 step before you have the stepper skip.)

3) Designing truss structures into printed parts can actually make your parts weaker. This area is counter intuitive to most. A "solid" arm can be both lighter and stronger than a part with holes cut in it to make it truss-like.
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

PreviousNext

Return to Wally

Who is online

Users browsing this forum: No registered users and 1 guest

cron