Design and Control of a Pantograph Robot - Revision history

Lynch: /* Mechanical Design */

2010-07-17T15:50:26Z

Mechanical Design

← Older revision		Revision as of 10:50, 17 July 2010
Line 45:		Line 45:
	=Mechanical Design=		=Mechanical Design=

	[[Image:PantographPhoto.jpg\|thumb\|200px\|[[A photo of the final design. Click on this caption to see a video.~~\|PantographVideo.mp4~~]]\|right]]		[[Image:PantographPhoto.jpg\|thumb\|200px\|[[Media:PantographVideo.mp4\|A photo of the final design. Click on this caption to see a video.]]\|right]]
	The mechanical design is essential in order for the user to be able to accurately control the pantograph robot. The most important factors in accurate control are friction and mechanical coupling between the motors and arms. We also made it our goal to minimize the vertical deflection at the end of the arms. In this section I will go over how the design aims to meet these goals, the issues I ran into and possible areas of improvement. At right you can see a picture of the final design; click on the caption to see a video.		The mechanical design is essential in order for the user to be able to accurately control the pantograph robot. The most important factors in accurate control are friction and mechanical coupling between the motors and arms. We also made it our goal to minimize the vertical deflection at the end of the arms. In this section I will go over how the design aims to meet these goals, the issues I ran into and possible areas of improvement. At right you can see a picture of the final design; click on the caption to see a video.
	<br clear=all>		<br clear=all>

Lynch: /* Mechanical Design */

2010-07-17T15:48:46Z

Mechanical Design

← Older revision		Revision as of 10:48, 17 July 2010
Line 45:		Line 45:
	=Mechanical Design=		=Mechanical Design=

	[[Image:PantographPhoto.jpg\|thumb\|200px\|A photo of the final design.\|right]]		[[Image:PantographPhoto.jpg\|thumb\|200px\|[[A photo of the final design. Click on this caption to see a video.\|PantographVideo.mp4]]\|right]]
	The mechanical design is essential in order for the user to be able to accurately control the pantograph robot. The most important factors in accurate control are friction and mechanical coupling between the motors and arms. We also made it ~~out~~ goal to minimize the vertical deflection at the end of the arms. In this section I will go over how the design aims to meet these goals, the issues I ran into and possible areas of improvement.		The mechanical design is essential in order for the user to be able to accurately control the pantograph robot. The most important factors in accurate control are friction and mechanical coupling between the motors and arms. We also made it our goal to minimize the vertical deflection at the end of the arms. In this section I will go over how the design aims to meet these goals, the issues I ran into and possible areas of improvement. At right you can see a picture of the final design; click on the caption to see a video.
	<br clear=all>		<br clear=all>

Lynch: /* Preliminary Results */

2010-07-17T15:39:34Z

Preliminary Results

← Older revision		Revision as of 10:39, 17 July 2010
Line 296:		Line 296:
	\|}		\|}
	<br clear=all><br>		<br clear=all><br>

			You can see a video of the pantograph robot [[Media:PantographVideo.mp4\|here]].

	=Next Steps=		=Next Steps=

Lynch: /* Mechanical Design */

2010-07-17T15:36:03Z

Mechanical Design

← Older revision		Revision as of 10:36, 17 July 2010
Line 49:		Line 49:
	<br clear=all>		<br clear=all>

	[[Image:PantographCAD.jpg\|thumb\|500px\|[[Full CAD Assembly of Pantograph Robot]]\|right]]		[[Image:PantographCAD.jpg\|thumb\|500px\|Full CAD Assembly of Pantograph Robot\|right]]
	The entire CAD assembly of the pantograph robot can be seen at the right. We can see a mounting piece used to attach the robot to an 80/20 frame is attached to the rear of the robot on either side of the "shoulders" (pictured in red). These "shoulder" are the base of the entire design. A mounting plate for the attachment of various size motors is attached to the top, below which two couplings are set in counter-bores within the "shoulder". These couplings attach the motors to the left and right "biceps", the upper part of the arms. These "biceps" are then attached to "forearms", the lower part of the arms. The "forearms" are then attached to the "biceps" (at the "elbow") and to each other (at the "wrist") with specially made pins.		The entire CAD assembly of the pantograph robot can be seen at the right. We can see a mounting piece used to attach the robot to an 80/20 frame is attached to the rear of the robot on either side of the "shoulders" (pictured in red). These "shoulder" are the base of the entire design. A mounting plate for the attachment of various size motors is attached to the top, below which two couplings are set in counter-bores within the "shoulder". These couplings attach the motors to the left and right "biceps", the upper part of the arms. These "biceps" are then attached to "forearms", the lower part of the arms. The "forearms" are then attached to the "biceps" (at the "elbow") and to each other (at the "wrist") with specially made pins.

Line 57:		Line 57:


	[[Image:Pantograph_ExpCAD.jpg\|thumb\|500px\|[[Exploded View of Pantograph Robot]]\|right]]		[[Image:Pantograph_ExpCAD.jpg\|thumb\|500px\|Exploded View of Pantograph Robot\|right]]
	We initially believed that friction would be the largest hurdle to overcome, therefore the base of the design is focused on reducing friction. We chose to use 26mm Thrust Bearings to support the load of the arms at the "shoulder" as well as reduce friction. To the right we can see an exploded view of the entire assembly. In this image we can see the thrust bearings (pictured in black) are sandwiched between the couplings and "shoulder" (on the top) and between a pair of washers and the "shoulder" (on the underside of the assembly). The "biceps" are then attached to the "shoulder" with a screw running through the center of the washer and bottom thrust bearing and into the bottom of the coupling. By tightening this screw the thrust bearings are firmly pressed between coupling and "shoulder" and "bicep" and "shoulder". This sandwiching of thrust bearings theoretically eliminates friction at the "shoulder" as well as vertical deflection at the end of the "biceps". However, in order for this to work properly, the coupling, bearings, and washers, must slide perfectly into the holes and counter-bores drilled for them in the "shoulder".		We initially believed that friction would be the largest hurdle to overcome, therefore the base of the design is focused on reducing friction. We chose to use 26mm Thrust Bearings to support the load of the arms at the "shoulder" as well as reduce friction. To the right we can see an exploded view of the entire assembly. In this image we can see the thrust bearings (pictured in black) are sandwiched between the couplings and "shoulder" (on the top) and between a pair of washers and the "shoulder" (on the underside of the assembly). The "biceps" are then attached to the "shoulder" with a screw running through the center of the washer and bottom thrust bearing and into the bottom of the coupling. By tightening this screw the thrust bearings are firmly pressed between coupling and "shoulder" and "bicep" and "shoulder". This sandwiching of thrust bearings theoretically eliminates friction at the "shoulder" as well as vertical deflection at the end of the "biceps". However, in order for this to work properly, the coupling, bearings, and washers, must slide perfectly into the holes and counter-bores drilled for them in the "shoulder".

Line 65:		Line 65:
	<br clear=all>		<br clear=all>

	[[Image:Pantograph_JointCAD.jpg\|thumb\|500px\|[[Close-ups of both the "Elbow" and "Wrist" Joints ]]\|right]]		[[Image:Pantograph_JointCAD.jpg\|thumb\|500px\|Close-ups of both the "Elbow" and "Wrist" Joints \|right]]
	Mechanical stops to prevent "hyper-extension" of the joints was another important design factor. You can see in the close up of the image to the right two small tabs. These tabs were left in the "forearm" design when machined. When the arm retracts, the corners of the opposing "forearm" hit these stops before the arm can "hyper-extend" inwards. Similarly, when machining the "biceps" from the aluminum U-channel, a little extra material was left in order to prevent "hyper-extension" of the "elbows".		Mechanical stops to prevent "hyper-extension" of the joints was another important design factor. You can see in the close up of the image to the right two small tabs. These tabs were left in the "forearm" design when machined. When the arm retracts, the corners of the opposing "forearm" hit these stops before the arm can "hyper-extend" inwards. Similarly, when machining the "biceps" from the aluminum U-channel, a little extra material was left in order to prevent "hyper-extension" of the "elbows".

	<br clear=all>		<br clear=all>

	[[Image:Pantograph_ClampCAD.jpg\|thumb\|500px\|[[Close-up of "Shoulder" Joint and Split-Clamp]]\|right]]		[[Image:Pantograph_ClampCAD.jpg\|thumb\|500px\|Close-up of "Shoulder" Joint and Split-Clamp\|right]]
	The final, and possibly the most important design consideration, came after it became clear that the originally designed solution would not work. It was originally believed that based on the theoretical load of the thrust bearings, we would be able to tighten down on the bolt which joins "bicep" to coupling, so much that the two would also be coupled radially. However, due to imperfections in prototyping (specifically, very minute differences in the angle of the surfaces which the thrust bearings rest on caused by imperfect mill alignment) we were unable to tighten down on this bolt much more than to keep the surfaces pressed together. As a solution to this problem we decided to modify the design of the upper "bicep", effectively turning it into a split-clamp. This new design can be seen in the image to the right.		The final, and possibly the most important design consideration, came after it became clear that the originally designed solution would not work. It was originally believed that based on the theoretical load of the thrust bearings, we would be able to tighten down on the bolt which joins "bicep" to coupling, so much that the two would also be coupled radially. However, due to imperfections in prototyping (specifically, very minute differences in the angle of the surfaces which the thrust bearings rest on caused by imperfect mill alignment) we were unable to tighten down on this bolt much more than to keep the surfaces pressed together. As a solution to this problem we decided to modify the design of the upper "bicep", effectively turning it into a split-clamp. This new design can be seen in the image to the right.
	The design change was achieved by drilling a larger hole in the "bicep" which the coupling could then slide into. Relief cuts were then made in both directions outward from this hole to allow for the deformation of the surrounding metal to pinch the end of the coupling in the hole. This strain was induced by attaching a small plate with a hole at the center to the open side of the U-channel and tapping a hole through the back of the channel. A bolt was then passed through the hole and tightened down until the split-clamp tightly gripped the end of the coupling.		The design change was achieved by drilling a larger hole in the "bicep" which the coupling could then slide into. Relief cuts were then made in both directions outward from this hole to allow for the deformation of the surrounding metal to pinch the end of the coupling in the hole. This strain was induced by attaching a small plate with a hole at the center to the open side of the U-channel and tapping a hole through the back of the channel. A bolt was then passed through the hole and tightened down until the split-clamp tightly gripped the end of the coupling.

Lynch: /* Mechanical Design */

2010-07-17T15:34:20Z

Mechanical Design

← Older revision		Revision as of 10:34, 17 July 2010
Line 44:		Line 44:

	=Mechanical Design=		=Mechanical Design=

			[[Image:PantographPhoto.jpg\|thumb\|200px\|A photo of the final design.\|right]]
	The mechanical design is essential in order for the user to be able to accurately control the pantograph robot. The most important factors in accurate control are friction and mechanical coupling between the motors and arms. We also made it out goal to minimize the vertical deflection at the end of the arms. In this section I will go over how the design aims to meet these goals, the issues I ran into and possible areas of improvement.		The mechanical design is essential in order for the user to be able to accurately control the pantograph robot. The most important factors in accurate control are friction and mechanical coupling between the motors and arms. We also made it out goal to minimize the vertical deflection at the end of the arms. In this section I will go over how the design aims to meet these goals, the issues I ran into and possible areas of improvement.
			<br clear=all>

	[[Image:PantographCAD.jpg\|thumb\|500px\|[[Full CAD Assembly of Pantograph Robot]]\|right]]		[[Image:PantographCAD.jpg\|thumb\|500px\|[[Full CAD Assembly of Pantograph Robot]]\|right]]

SeanWood: /* Issues and Improvements */

2010-07-09T16:43:19Z

Issues and Improvements

← Older revision		Revision as of 11:43, 9 July 2010
Line 85:		Line 85:
	The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.		The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.

	All of the CAD files (Solidworks) can be found Here:[[~~Pantograph CAD~~]].		All of the CAD files (Solidworks) can be found Here:[[Image:PantographRobot.zip]].

	=Circuit Design=		=Circuit Design=

SeanWood: /* Issues and Improvements */

2010-07-09T16:30:36Z

Issues and Improvements

← Older revision		Revision as of 11:30, 9 July 2010
Line 85:		Line 85:
	The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.		The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.

	All of the CAD files (Solidworks) can be found [[~~Here~~]].		All of the CAD files (Solidworks) can be found Here:[[Pantograph CAD]].

	=Circuit Design=		=Circuit Design=

SeanWood: /* Issues and Improvements */

2010-07-09T16:29:31Z

Issues and Improvements

← Older revision		Revision as of 11:29, 9 July 2010
Line 84:		Line 84:
	The simplest of these sources to fix is reinforcing the support plate, however, I do not think this is the most important factor in the problem. The only solution to the second problem is remaking the "biceps" and drilling and reaming the hole to precisely the right diameter, maybe even a thousandth or two undersized. The best solution would be to redesign the biceps and coupling completely, allowing for more contact area and therefore more frictional force during rotation. A slight modification that may would though would be to fillet the top and bottom of the split clamp hole turning the cross section of the edge into a triangle. Then turning an identical size and shape triangular cross section into the end of the coupling would cause the new mate to have an increased surface area, theoretically increasing the maximum torque of the clamp.		The simplest of these sources to fix is reinforcing the support plate, however, I do not think this is the most important factor in the problem. The only solution to the second problem is remaking the "biceps" and drilling and reaming the hole to precisely the right diameter, maybe even a thousandth or two undersized. The best solution would be to redesign the biceps and coupling completely, allowing for more contact area and therefore more frictional force during rotation. A slight modification that may would though would be to fillet the top and bottom of the split clamp hole turning the cross section of the edge into a triangle. Then turning an identical size and shape triangular cross section into the end of the coupling would cause the new mate to have an increased surface area, theoretically increasing the maximum torque of the clamp.
	The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.		The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.

			All of the CAD files (Solidworks) can be found [[Here]].

	=Circuit Design=		=Circuit Design=

SeanWood: /* Mechanical Design */

2010-07-09T16:26:30Z

Mechanical Design

← Older revision		Revision as of 11:26, 9 July 2010
Line 67:		Line 67:
	<br clear=all>		<br clear=all>

	[[Image:Pantograph_ClampCAD.jpg\|thumb\|500px\|[[~~Exploded View~~ of ~~Pantograph~~ ~~Robot~~]]\|right]]		[[Image:Pantograph_ClampCAD.jpg\|thumb\|500px\|[[Close-up of "Shoulder" Joint and Split-Clamp]]\|right]]
	The final, and possibly the most important design consideration, came after it became clear that the originally designed solution would not work. It was originally believed that based on the theoretical load of the thrust bearings, we would be able to tighten down on the bolt which joins "bicep" to coupling, so much that the two would also be coupled radially. However, due to imperfections in prototyping (specifically, very minute differences in the angle of the surfaces which the thrust bearings rest on caused by imperfect mill alignment) we were unable to tighten down on this bolt much more than to keep the surfaces pressed together. As a solution to this problem we decided to modify the design of the upper "bicep", effectively turning it into a split-clamp. This new design can be seen in the image to the right.		The final, and possibly the most important design consideration, came after it became clear that the originally designed solution would not work. It was originally believed that based on the theoretical load of the thrust bearings, we would be able to tighten down on the bolt which joins "bicep" to coupling, so much that the two would also be coupled radially. However, due to imperfections in prototyping (specifically, very minute differences in the angle of the surfaces which the thrust bearings rest on caused by imperfect mill alignment) we were unable to tighten down on this bolt much more than to keep the surfaces pressed together. As a solution to this problem we decided to modify the design of the upper "bicep", effectively turning it into a split-clamp. This new design can be seen in the image to the right.
	The design change was achieved by drilling a larger hole in the "bicep" which the coupling could then slide into. Relief cuts were then made in both directions outward from this hole to allow for the deformation of the surrounding metal to pinch the end of the coupling in the hole. This strain was induced by attaching a small plate with a hole at the center to the open side of the U-channel and tapping a hole through the back of the channel. A bolt was then passed through the hole and tightened down until the split-clamp tightly gripped the end of the coupling.		The design change was achieved by drilling a larger hole in the "bicep" which the coupling could then slide into. Relief cuts were then made in both directions outward from this hole to allow for the deformation of the surrounding metal to pinch the end of the coupling in the hole. This strain was induced by attaching a small plate with a hole at the center to the open side of the U-channel and tapping a hole through the back of the channel. A bolt was then passed through the hole and tightened down until the split-clamp tightly gripped the end of the coupling.

SeanWood: /* Issues and Improvements */

2010-07-09T16:25:26Z

Issues and Improvements

← Older revision		Revision as of 11:25, 9 July 2010
Line 76:		Line 76:

	There were a handful of problems I ran into while machining and assembling the pantograph robot.		There were a handful of problems I ran into while machining and assembling the pantograph robot.

	As I mentioned above there was the initial issue of not being able to tighten down the shoulder/coupling/bicep bolt as much as I calculated would be possible due to machining defects. This slight error was not caused by burrs or chips under the part during the machining process but misalignment of the machine itself. You must always keep in mind that your part can only be as accurate as the machine creating it, and mills (particularly the ones down in the shop which are often misused) are rarely aligned perfectly. This very slight difference between the angles of the two thrust bearings caused friction at one end of rotation under high loads. We got around this issue by lightening the load and introducing the split-clamp design to couple the two pieces radially. In future versions closer attention should be paid to the precise alignment of the mill creating the part.		As I mentioned above there was the initial issue of not being able to tighten down the shoulder/coupling/bicep bolt as much as I calculated would be possible due to machining defects. This slight error was not caused by burrs or chips under the part during the machining process but misalignment of the machine itself. You must always keep in mind that your part can only be as accurate as the machine creating it, and mills (particularly the ones down in the shop which are often misused) are rarely aligned perfectly. This very slight difference between the angles of the two thrust bearings caused friction at one end of rotation under high loads. We got around this issue by lightening the load and introducing the split-clamp design to couple the two pieces radially. In future versions closer attention should be paid to the precise alignment of the mill creating the part.

	Another critical friction issue was one at the wrist pin. When the arms were brought to full extension there was a noticeable increase in friction at the wrist. Upon further investigation it was found that this was caused by a very slight misalignment between the central axis of the bearings in either "forearm". This misalignment, stemming from back at the shoulder, caused the wrist-pin to press tightly against the walls of the bearing. We worked around this issue by relieving slightly the tolerance at the "elbow" joint, allowing for a very slight misalignment. In future designs a single self-aligning ball bearing could be used as it allows for several degrees of misalignment (Part Number:6660k11 found at [http://www.mcmaster.com/#6660k11/=7vxwtc McMaster-Carr] would most likely work within the constraints of the current design).		Another critical friction issue was one at the wrist pin. When the arms were brought to full extension there was a noticeable increase in friction at the wrist. Upon further investigation it was found that this was caused by a very slight misalignment between the central axis of the bearings in either "forearm". This misalignment, stemming from back at the shoulder, caused the wrist-pin to press tightly against the walls of the bearing. We worked around this issue by relieving slightly the tolerance at the "elbow" joint, allowing for a very slight misalignment. In future designs a single self-aligning ball bearing could be used as it allows for several degrees of misalignment (Part Number:6660k11 found at [http://www.mcmaster.com/#6660k11/=7vxwtc McMaster-Carr] would most likely work within the constraints of the current design).

			The most critical issue which still remains though is the slipping of the shoulder split-clamp at high speeds. There are a handful of reasons why this might be: the support plate for the bolt is made of relatively thin aluminum and cannot support a large load, the hole drilled for the end of the coupling is a couple thousandths larger than the coupling diameter itself as opposed to being a press fit, the surface contact area between clamp and coupling is too small.
			The simplest of these sources to fix is reinforcing the support plate, however, I do not think this is the most important factor in the problem. The only solution to the second problem is remaking the "biceps" and drilling and reaming the hole to precisely the right diameter, maybe even a thousandth or two undersized. The best solution would be to redesign the biceps and coupling completely, allowing for more contact area and therefore more frictional force during rotation. A slight modification that may would though would be to fillet the top and bottom of the split clamp hole turning the cross section of the edge into a triangle. Then turning an identical size and shape triangular cross section into the end of the coupling would cause the new mate to have an increased surface area, theoretically increasing the maximum torque of the clamp.
			The best solution though would be to purchase a U-channel with a thicker wall thickness, and machine out the inside of the walls everywhere except at the "shoulder" end of the "bicep". This would allow for the machinist to more precisely drill the hole, and for increased clamping force and surface area.

	=Circuit Design=		=Circuit Design=