23rd Century CNC
Plastic
- Joined
- Dec 2, 2018
- Location
- Kinsman, OH, USA
I recently purchased a vacuum plate system and spent a lot of time on this forum and the manufacturers website learning about how to best use it. I learned a lot but there were still many questions I had to answer myself. This post will hopefully contribute a few more pieces of information that someone else learning vacuum workholding might benefit from.
Many other sources already detail how vacuum workholding actually works so I'll just state the basics here. The holding power of a vacuum system is directly related to surface area of vacuum on your part. The relationship is surface area multiplied by atmospheric pressure. If you were holding a 10" square part then you'd have 100in^2 of surface area under vacuum. Multiply this by atmospheric pressure, approx 14psi, and you'd have 1400 pounds of down-force holding your part. There will always be some inefficiency from the sealing gasket, and quality of the part surface against the gasket. Correct me if I'm wrong, it doesn't matter what vacuum pump/generator you use so long as it's flow rate exceeds the leakage rate in your system.
I purchased a Pierson Workholding system.
SmartVac II Starter Kit SmartVac II Starter Package - 9.5" x 14" Base
It included the 9.5" x 14" base plate and the Venturi vacuum generator.
I chose this model because it was narrow enough to allow holding it in standard 6" vises with the jaws reversed.
View attachment 243536
As a job shop I like to keep my two vises on the mill as much as possible. The 9.5" x 14" size is large enough to accommodate most jobs we'd run on this machine. And if we need a larger plate in the future then we'll be more comfortable spending the extra money now having used a vacuum system.
The system itself is pretty well defined on the manufacturers website using photos, videos, and the provided installation documentation. The only tools I really needed were some scissors and a T-Nut.
After gripping the base in my vises I quickly verified that it didn't bow very much from the vise pressure, the plate itself is Blanchard ground and has peripheral slots that allow for bolting down to a T-slot table, this should keep the plate flatter than holding it in the vise but I was still able to keep it within 0.002" flatness over the whole plate.
The base plate has a grid pattern of gasket grooves, I've ran a couple parts directly on this grid, but everything else I have done has been using their sacrificial top plates.
Additional to the starter kit I also purchased a matching Top Plate and their Gasket Slotting End Mill.
SmartVac II Top Plate - 9.5" x 14"
SmartVac II "Gasket Slot" Endmill
The top plates look to be aluminum tooling plate with four accurate holes in the corners. There are four shoulder screws provided in the starter kit which locate the top plate onto the base plate using those screws.
The top plates allow you to customize your gasket groove pattern as well as serve as sacrificial components allowing you to drill through or add locating features. After bolting on a top plate I noticed no significant deviation from the base plates flatness. There is a video on Pierson's website showing their preferred method of machining your top plate. They show adding gasket groove all around the outer groove in the base plate then bolting on the top plate. They advise you to machine your gasket groove and any blind features with the vacuum generator running, then to release the vacuum before doing any through features such as the drilled hole(s) for porting the vacuum up to your gasketed area. Pierson also recommends drilling through using a standard tap drill size such that you can plug the hole later if you wanted to. I believe they recommended the tap drill for a #10 screw, but I chose instead to use M3. I am not sure but I do not think this would significantly affect air flow rates.
Pierson sell a custom 0.118" end mill for cutting gasket grooves (for their 1/8" gasket cord). I used this end mill at 6000 RPM, 18 IPM, with a 2 degree ramp entry, and a depth of 0.095". This cut the slots great and left the designed chamfer around the top edges. Pierson do also mention the potential benefits of cutting shallower "distribution" channels to help air flow from the extremities of your gasketed area to the vacuum port hole. I don't know if these channels have much effect on vacuum performance but I added them anyway since I can't think of a detrimental effect. For these distribution channels I used the same end mill and just slotted 0.040" deep then used a 1/8" chamfer tool to break the edges, having any burrs under your part wouldn't help your sealing efficiency.
Note that the vacuum gasket cord has a round cross section. However once you load the plate and turn on the vacuum this 0.125" cord gets squished down into the 0.095" slot and forms a square cross section. This drastically increases the sealing surface area of the gasket. I therefore think the depth of slotting for the gasket groove is fairly important. By following Pierson's directions for inserting the cord into the groove I had no issues with sealing, cut the ends square and push into the groove to create compression.
View attachment 243535
I have only used 1/8" gasket though I have also purchased an assortment of other sizes from Pierson. The 1/16" gasket might be worth experimenting with if you have lots of drilled holes through your part, as the 1/8" gasket begins to take away a lot of your vacuum surface area if there are many through holes to seal. Also gasket has a minimum bend radius, therefore limiting how precisely you can seal your part with through features, see the photo below where the gasket has to snake around features where it can't bend tight enough to enter. I was able to seal a small through hole with an inner radius of 0.1", i.e. the ID of the gasket groove was 0.200" and the OD was 0.436". This seemed really sketchy to me but it worked through half a dozen parts, each with 7hrs of cycle time. So I don't know if you could get away with a tighter bend radius, but I would recommend stopping at R0.125" while using the 1/8" gasket.
View attachment 243533
In actual use of the vacuum system I found the Venturi to be reliable and tolerant of coolant as advertised. Many years ago we experimented with a vacuum pump and found out that you had to be cautious of getting coolant into the pump. The Pierson Venturi unit just gets clamped into a T-slot and you run push-connect air tubing between the vacuum base, the Venturi, a pressure regulator, and a air supply. On our Haas machine there are spare air ports on the side of the machine. All hardware was included in the starter kit, I cut the tubing to the lengths required for the travels of my machine and installed everything. The supplied pressure regulator comes with a magnet glued to it for mounting on sheet metal. I stuck this right to my machine cabinet and didn't have any issues with vibration. I did not have to drill any holes or do any machine modifications to install this system.
View attachment 243534
When I ran my first part it was a 3/8" thick Delrin strip, about as long as my 14" plate and a few inches wide. I had difficulty holding this part without it shifting. I realized I needed more surface area for the vacuum. I was able to modify my top plate pattern to get a few more square inches of area as well as extending the gasket out to every corner of the part. I believe the end mill created enough lifting force in the corners of the part to break the vacuum and shift the part. So getting the gasket as close to the edge while still leaving some buffer distance is critical. I also noticed that cutting on the shorter 3" sides was much more likely to shift the part than when cutting on the long sides. I mitigated this by a combination of methods. I switched to a smaller diameter tool and isolated my cuts to single edges rather than cutting around corners, I had the most trouble with lift when the end mill rounded an external corner. I don't know for sure if this next decision made a difference or not, but after a couple trial and errors I switched my coolant off, I just had it on by default and didn't need it for this part anyway. I suspected that the face milled bottom side of this Delrin strip may have been sliding due to the coolant too.
The next part I cut was a 1" thick 10" square slab of aluminum tooling plate. This part would end up as a "W" shaped part. To hold this on my smaller plate I actually rotated it 45 degrees, such that the section that now overhung my plate was all getting milled away and all the usable vacuum area was contained within my 9.5" vacuum plate. I didn't want to throw this part if it lifted cutting that overhung area so I did something different. There were a couple through holes that happened to be nominal inch sizes like 1/4" and 3/8". I added a couple interpolated holes to my Top Plate. Then while running the part I initially drilled those holes and dropped dowel pins through the part and into the Top Plate (sealed with gasket and also not drilled entirely through the top plate). These dowel pins allowed me to go much more aggressively after this part and mill it in a reasonable time as compared to how we would have usually done it on a sub-plate with toe clamps. If I had a spare piece of material I would have bumped up my feed rates and widths of cut, but I easily was able to use a 1/2" mill at full depth profiling the part at 120 IPM and 0.050" WOC. The two dowel pins seemed to do the trick.
With the dowel pin trick I now felt much more comfortable taking on some work specifically for the vacuum plate, the kind of work which would have been tricky and extremely tedious to do with clamps or tape. These parts were aluminum plates that each were a few inches wide by 12" long, 1/4" starting thickness with nearly the whole parts pocketed out to a floor and wall thickness of 0.040" and completely surrounded by rings of drilled and tapped holes. Each part required about 3.5hrs of pocketing and o-ring groove slotting, with a +/-0.002" tolerance on may pocket sizes and locations. These parts could not be allowed to shift at all, and luckily they had a couple dowel pin holes in them. I initially faced the parts on both sides using the vacuum plate. Then the first features I added were these dowel holes. I had those dowel locations reamed in the Top Plate, allowing me to drop in a couple tightly fitting pins to keep the part from shifting. This worked excellent and I had no issues with movement. But one plate style had through holes while the other had the associated tapped holes. The plate with through holes therefore had significantly less surface area. Then I learned another trick, nesting parts. I was able to fit one of each part on the same top plate. This let me buy fewer, larger sheets of material. And run both parts simultaneously as a set. And aid the part with less surface area by having it on the same setup as the part with more surface area, this lessened the chance of the far corners lifting and combined with the dowel pins produced a very secure setup.
So hopefully the information above helps someone else get into vacuum workholding, feel free to reach out with questions as I'm sure I'm forgetting some things I learned.
In summary...
The Venturi "SmartVac II" system from Pierson works great.
I have only used 1/8" gasket with their top plates and their custom end mill, but I will probably try the 1/16" gasket at some point and probably use the same ration of depth and width as for the 1/8" stuff.
Maximize all usable surface area to get the most holding force and prevent lifting when cutting around corners.
When milling external corners...either don't...or at least reduce feed rates.
Coolant may have an effect on holding ability.
Vacuum gasket has a minimum bend radius, particularly when sealing drilled holes, right angles are doable for other areas but they do cause "bunching" of the gasket at that angle's corner.
Unknown whether distribution channels have an effect on performance
Only use the largest tool diameters as needed to keep lifting forces as low as possible.
Dowel pins or other "locational" references dramatically increase rigidity, they do not prevent lift but do eliminate shifting. Maybe adding a bolt partway through a cycle would be the best method since it'd eliminate lift and shift. But distant areas could still lift.
Nest parts to increase holding power, leave thin webs between parts to keep them connected as a whole sheet. Design webs to be easily removed manually after machining.
Many other sources already detail how vacuum workholding actually works so I'll just state the basics here. The holding power of a vacuum system is directly related to surface area of vacuum on your part. The relationship is surface area multiplied by atmospheric pressure. If you were holding a 10" square part then you'd have 100in^2 of surface area under vacuum. Multiply this by atmospheric pressure, approx 14psi, and you'd have 1400 pounds of down-force holding your part. There will always be some inefficiency from the sealing gasket, and quality of the part surface against the gasket. Correct me if I'm wrong, it doesn't matter what vacuum pump/generator you use so long as it's flow rate exceeds the leakage rate in your system.
I purchased a Pierson Workholding system.
SmartVac II Starter Kit SmartVac II Starter Package - 9.5" x 14" Base
It included the 9.5" x 14" base plate and the Venturi vacuum generator.
I chose this model because it was narrow enough to allow holding it in standard 6" vises with the jaws reversed.
View attachment 243536
As a job shop I like to keep my two vises on the mill as much as possible. The 9.5" x 14" size is large enough to accommodate most jobs we'd run on this machine. And if we need a larger plate in the future then we'll be more comfortable spending the extra money now having used a vacuum system.
The system itself is pretty well defined on the manufacturers website using photos, videos, and the provided installation documentation. The only tools I really needed were some scissors and a T-Nut.
After gripping the base in my vises I quickly verified that it didn't bow very much from the vise pressure, the plate itself is Blanchard ground and has peripheral slots that allow for bolting down to a T-slot table, this should keep the plate flatter than holding it in the vise but I was still able to keep it within 0.002" flatness over the whole plate.
The base plate has a grid pattern of gasket grooves, I've ran a couple parts directly on this grid, but everything else I have done has been using their sacrificial top plates.
Additional to the starter kit I also purchased a matching Top Plate and their Gasket Slotting End Mill.
SmartVac II Top Plate - 9.5" x 14"
SmartVac II "Gasket Slot" Endmill
The top plates look to be aluminum tooling plate with four accurate holes in the corners. There are four shoulder screws provided in the starter kit which locate the top plate onto the base plate using those screws.
The top plates allow you to customize your gasket groove pattern as well as serve as sacrificial components allowing you to drill through or add locating features. After bolting on a top plate I noticed no significant deviation from the base plates flatness. There is a video on Pierson's website showing their preferred method of machining your top plate. They show adding gasket groove all around the outer groove in the base plate then bolting on the top plate. They advise you to machine your gasket groove and any blind features with the vacuum generator running, then to release the vacuum before doing any through features such as the drilled hole(s) for porting the vacuum up to your gasketed area. Pierson also recommends drilling through using a standard tap drill size such that you can plug the hole later if you wanted to. I believe they recommended the tap drill for a #10 screw, but I chose instead to use M3. I am not sure but I do not think this would significantly affect air flow rates.
Pierson sell a custom 0.118" end mill for cutting gasket grooves (for their 1/8" gasket cord). I used this end mill at 6000 RPM, 18 IPM, with a 2 degree ramp entry, and a depth of 0.095". This cut the slots great and left the designed chamfer around the top edges. Pierson do also mention the potential benefits of cutting shallower "distribution" channels to help air flow from the extremities of your gasketed area to the vacuum port hole. I don't know if these channels have much effect on vacuum performance but I added them anyway since I can't think of a detrimental effect. For these distribution channels I used the same end mill and just slotted 0.040" deep then used a 1/8" chamfer tool to break the edges, having any burrs under your part wouldn't help your sealing efficiency.
Note that the vacuum gasket cord has a round cross section. However once you load the plate and turn on the vacuum this 0.125" cord gets squished down into the 0.095" slot and forms a square cross section. This drastically increases the sealing surface area of the gasket. I therefore think the depth of slotting for the gasket groove is fairly important. By following Pierson's directions for inserting the cord into the groove I had no issues with sealing, cut the ends square and push into the groove to create compression.
View attachment 243535
I have only used 1/8" gasket though I have also purchased an assortment of other sizes from Pierson. The 1/16" gasket might be worth experimenting with if you have lots of drilled holes through your part, as the 1/8" gasket begins to take away a lot of your vacuum surface area if there are many through holes to seal. Also gasket has a minimum bend radius, therefore limiting how precisely you can seal your part with through features, see the photo below where the gasket has to snake around features where it can't bend tight enough to enter. I was able to seal a small through hole with an inner radius of 0.1", i.e. the ID of the gasket groove was 0.200" and the OD was 0.436". This seemed really sketchy to me but it worked through half a dozen parts, each with 7hrs of cycle time. So I don't know if you could get away with a tighter bend radius, but I would recommend stopping at R0.125" while using the 1/8" gasket.
View attachment 243533
In actual use of the vacuum system I found the Venturi to be reliable and tolerant of coolant as advertised. Many years ago we experimented with a vacuum pump and found out that you had to be cautious of getting coolant into the pump. The Pierson Venturi unit just gets clamped into a T-slot and you run push-connect air tubing between the vacuum base, the Venturi, a pressure regulator, and a air supply. On our Haas machine there are spare air ports on the side of the machine. All hardware was included in the starter kit, I cut the tubing to the lengths required for the travels of my machine and installed everything. The supplied pressure regulator comes with a magnet glued to it for mounting on sheet metal. I stuck this right to my machine cabinet and didn't have any issues with vibration. I did not have to drill any holes or do any machine modifications to install this system.
View attachment 243534
When I ran my first part it was a 3/8" thick Delrin strip, about as long as my 14" plate and a few inches wide. I had difficulty holding this part without it shifting. I realized I needed more surface area for the vacuum. I was able to modify my top plate pattern to get a few more square inches of area as well as extending the gasket out to every corner of the part. I believe the end mill created enough lifting force in the corners of the part to break the vacuum and shift the part. So getting the gasket as close to the edge while still leaving some buffer distance is critical. I also noticed that cutting on the shorter 3" sides was much more likely to shift the part than when cutting on the long sides. I mitigated this by a combination of methods. I switched to a smaller diameter tool and isolated my cuts to single edges rather than cutting around corners, I had the most trouble with lift when the end mill rounded an external corner. I don't know for sure if this next decision made a difference or not, but after a couple trial and errors I switched my coolant off, I just had it on by default and didn't need it for this part anyway. I suspected that the face milled bottom side of this Delrin strip may have been sliding due to the coolant too.
The next part I cut was a 1" thick 10" square slab of aluminum tooling plate. This part would end up as a "W" shaped part. To hold this on my smaller plate I actually rotated it 45 degrees, such that the section that now overhung my plate was all getting milled away and all the usable vacuum area was contained within my 9.5" vacuum plate. I didn't want to throw this part if it lifted cutting that overhung area so I did something different. There were a couple through holes that happened to be nominal inch sizes like 1/4" and 3/8". I added a couple interpolated holes to my Top Plate. Then while running the part I initially drilled those holes and dropped dowel pins through the part and into the Top Plate (sealed with gasket and also not drilled entirely through the top plate). These dowel pins allowed me to go much more aggressively after this part and mill it in a reasonable time as compared to how we would have usually done it on a sub-plate with toe clamps. If I had a spare piece of material I would have bumped up my feed rates and widths of cut, but I easily was able to use a 1/2" mill at full depth profiling the part at 120 IPM and 0.050" WOC. The two dowel pins seemed to do the trick.
With the dowel pin trick I now felt much more comfortable taking on some work specifically for the vacuum plate, the kind of work which would have been tricky and extremely tedious to do with clamps or tape. These parts were aluminum plates that each were a few inches wide by 12" long, 1/4" starting thickness with nearly the whole parts pocketed out to a floor and wall thickness of 0.040" and completely surrounded by rings of drilled and tapped holes. Each part required about 3.5hrs of pocketing and o-ring groove slotting, with a +/-0.002" tolerance on may pocket sizes and locations. These parts could not be allowed to shift at all, and luckily they had a couple dowel pin holes in them. I initially faced the parts on both sides using the vacuum plate. Then the first features I added were these dowel holes. I had those dowel locations reamed in the Top Plate, allowing me to drop in a couple tightly fitting pins to keep the part from shifting. This worked excellent and I had no issues with movement. But one plate style had through holes while the other had the associated tapped holes. The plate with through holes therefore had significantly less surface area. Then I learned another trick, nesting parts. I was able to fit one of each part on the same top plate. This let me buy fewer, larger sheets of material. And run both parts simultaneously as a set. And aid the part with less surface area by having it on the same setup as the part with more surface area, this lessened the chance of the far corners lifting and combined with the dowel pins produced a very secure setup.
So hopefully the information above helps someone else get into vacuum workholding, feel free to reach out with questions as I'm sure I'm forgetting some things I learned.
In summary...
The Venturi "SmartVac II" system from Pierson works great.
I have only used 1/8" gasket with their top plates and their custom end mill, but I will probably try the 1/16" gasket at some point and probably use the same ration of depth and width as for the 1/8" stuff.
Maximize all usable surface area to get the most holding force and prevent lifting when cutting around corners.
When milling external corners...either don't...or at least reduce feed rates.
Coolant may have an effect on holding ability.
Vacuum gasket has a minimum bend radius, particularly when sealing drilled holes, right angles are doable for other areas but they do cause "bunching" of the gasket at that angle's corner.
Unknown whether distribution channels have an effect on performance
Only use the largest tool diameters as needed to keep lifting forces as low as possible.
Dowel pins or other "locational" references dramatically increase rigidity, they do not prevent lift but do eliminate shifting. Maybe adding a bolt partway through a cycle would be the best method since it'd eliminate lift and shift. But distant areas could still lift.
Nest parts to increase holding power, leave thin webs between parts to keep them connected as a whole sheet. Design webs to be easily removed manually after machining.