A2 Wrote:
-------------------------------------------------------
> Guizmo, Nicholas Seward:
>
> I like the combination of the zero back lash of
> the Double Lamina Compliant Joint (DLCJ)
> bearing/joint, and the non linear toothless
> elliptical gear or cam correction as it affords
> weight, and part count reduction, moreover it
> works with the Simpson firm ware without further
> modifications!
>
> I believe I have a solution (or a starting point
> for discussion) that incorporates both the DLCJ
> bearing/joint, and the elliptical gear/cam, please
> see attached blueprints.
>
>
> Comments/Notes/Questions:
>
> I'm using wide control arms/linkages to help
> increase rigidity (side-to-side sway), and I have
> eliminated the tie rods that connect adjacent DLCJ
> bearings. In addition, I added DLCJ bearings to
> the upper shoulder pivot (stepper motor axis), and
> to the wrist joint (end effector).
>
> For reference, the DLCJ bearings are 1 inch wide
> (not shown), the DLCJ bearings center to center
> distance is 2.5 inches, and with an over all width
> of 3.75 inches. That's worst case, I have not
> taken advantage of all the surface area for
> tethering/binding.
>
> The elbow pivot was given enough clearance for a 4
> inch diameter pulley, for a 4:1 ratio from the
> stepper motor pulley. For reference, I have drawn
> in a 2 inch diameter timing pulley in the forearm
> joint.
>
> The arms are 6 inches long measured from the
> distal pivot of the arm, to the furthest DLCJ
> pivot, and the elliptical toothless gears are for
> reference only. The longest part is 6 inches.
>
> To adjust the timing belt tension the upper arm
> needs to be constructed of two parts (not shown).
> I'm considering guides of some flavor, plus dual
> turnbuckles or jack-screws.
>
> Can you utilize the DLCJ bearing on the pivotal
> base/pedestal (vertical axis of rotation)?
>
> Can the Simpson firmware be used in it's current
> configuration without modifications for the power
> train that I have drawn?
>
> Is the elliptical gear or cam profile scalable, or
> does the profile need to be recalculated for a
> given arm length?
>
> What plastic material will be used most often with
> this design? I would like to get a rough idea of
> what the wall thickness should be. I'm also
> thinking of shelling the part, and adding ribs in
> critical areas, instead of using 20% infilling.
> Will the shrink rate and warping be acceptable
> with no infilling, or does the infilling help
> prevent warping?
>
> I appreciate your feedback, and I'm looking
> forward to a geometric solution.
>
> A2
Hello A2!
I'm glad you're interested in this design, as I also believe part reduction is a very good goal for any reprap, and specially for an already good design like Simpson (which I also believe is the most reprap-able machine at this time).
I have a few comments on your design:
For what I see, you are trying to use the DLCJ as a replacement to bearings, which is good. I imagine the first joint (closer to the motor) will work like a knee, and that is a brilliant idea, because the motor mount now can be in the same axis of rotation and is simplified a lot. But the driven joint wouldn't work in that configuration,because the center of rotation is located on the center of the first arm end, so, that is the position the second "cam" has to be, and that's all, I belive it could work fine [Note: Not true, the center of rotation moves along the line of simmetry and its distance from the contact point tends to infinity as the angle betwen arms is closer to zero] . I draw a mockup:
[attachment 17579 brazo3b.jpg]
![]()
Please note that the centers of rotation are of the "more fixed" side of the DLCJ, so you have to measure the distances correctly. (Please correct me if I'm wrong) [I'm wrong, anyway, I believe something useful can be obtained from A2's idea of DLCJ instead of bearings. Let's make it work]
Regarding the lenght of arms, that is important. Well, it is important that both are the same lenght, because if they are different, the cams have to be different too. Anyway, it doesn't matter the lenght nor the size of the cams, the only thing important about the cams is that they mantain the ratio they are designed to mantain (which is variable). So, yes, the cams are scalable.
Regarding firmware and plastic I don't know. Can you believe I don't have a 3d printer? That is why I'm so interested in this design to be cheap and reliable. I wan t one!!!
Edit: Guys, I'm getting really excited!!
-------------------------------------------------------
> Guizmo, Nicholas Seward:
>
> I like the combination of the zero back lash of
> the Double Lamina Compliant Joint (DLCJ)
> bearing/joint, and the non linear toothless
> elliptical gear or cam correction as it affords
> weight, and part count reduction, moreover it
> works with the Simpson firm ware without further
> modifications!
>
> I believe I have a solution (or a starting point
> for discussion) that incorporates both the DLCJ
> bearing/joint, and the elliptical gear/cam, please
> see attached blueprints.
>
>
> Comments/Notes/Questions:
>
> I'm using wide control arms/linkages to help
> increase rigidity (side-to-side sway), and I have
> eliminated the tie rods that connect adjacent DLCJ
> bearings. In addition, I added DLCJ bearings to
> the upper shoulder pivot (stepper motor axis), and
> to the wrist joint (end effector).
>
> For reference, the DLCJ bearings are 1 inch wide
> (not shown), the DLCJ bearings center to center
> distance is 2.5 inches, and with an over all width
> of 3.75 inches. That's worst case, I have not
> taken advantage of all the surface area for
> tethering/binding.
>
> The elbow pivot was given enough clearance for a 4
> inch diameter pulley, for a 4:1 ratio from the
> stepper motor pulley. For reference, I have drawn
> in a 2 inch diameter timing pulley in the forearm
> joint.
>
> The arms are 6 inches long measured from the
> distal pivot of the arm, to the furthest DLCJ
> pivot, and the elliptical toothless gears are for
> reference only. The longest part is 6 inches.
>
> To adjust the timing belt tension the upper arm
> needs to be constructed of two parts (not shown).
> I'm considering guides of some flavor, plus dual
> turnbuckles or jack-screws.
>
> Can you utilize the DLCJ bearing on the pivotal
> base/pedestal (vertical axis of rotation)?
>
> Can the Simpson firmware be used in it's current
> configuration without modifications for the power
> train that I have drawn?
>
> Is the elliptical gear or cam profile scalable, or
> does the profile need to be recalculated for a
> given arm length?
>
> What plastic material will be used most often with
> this design? I would like to get a rough idea of
> what the wall thickness should be. I'm also
> thinking of shelling the part, and adding ribs in
> critical areas, instead of using 20% infilling.
> Will the shrink rate and warping be acceptable
> with no infilling, or does the infilling help
> prevent warping?
>
> I appreciate your feedback, and I'm looking
> forward to a geometric solution.
>
> A2
Hello A2!
I'm glad you're interested in this design, as I also believe part reduction is a very good goal for any reprap, and specially for an already good design like Simpson (which I also believe is the most reprap-able machine at this time).
I have a few comments on your design:
For what I see, you are trying to use the DLCJ as a replacement to bearings, which is good. I imagine the first joint (closer to the motor) will work like a knee, and that is a brilliant idea, because the motor mount now can be in the same axis of rotation and is simplified a lot. But the driven joint wouldn't work in that configuration,
[attachment 17579 brazo3b.jpg]
Regarding the lenght of arms, that is important. Well, it is important that both are the same lenght, because if they are different, the cams have to be different too. Anyway, it doesn't matter the lenght nor the size of the cams, the only thing important about the cams is that they mantain the ratio they are designed to mantain (which is variable). So, yes, the cams are scalable.
Regarding firmware and plastic I don't know. Can you believe I don't have a 3d printer? That is why I'm so interested in this design to be cheap and reliable. I wan t one!!!
Edit: Guys, I'm getting really excited!!