Understanding Vacuum Hold-Down
There is much confusion about what kind of vacuum pumps are needed for different applications regarding hold-down on CNC Routers. Many of us think that more horsepower automatically results in better hold-down and opt for more HP and of course more electric bills. I have used everything from reverse blowers from Grainger that will work on 220 single phase to 25 HP regenerative blowers to positive displacement pumps such as Becker, Busch, or others. I have built CNC routers for 13 years and just lately feel like I understand it enough to really help someone else understand it. For most people including myself, we may think that a pump holds down parts because it "sucks the part down with air". When we understand what really holds the part it will help us understand more about holding down parts. Piab is a company that makes small vacuum cups, small pumps, etc. Several years ago I read a book they had and it explains very well how it works and it was very interesting. Basically without getting too deep into science this is it.
On what is called a standard day which is 59 degrees F at sea level air weighs about 15 lbs per square inch. Think of it as a column of air 1 inch square that goes all the way toward outer space until there are no air molecules. If you live in the mountains say at 2000 feet in elevation then your column of air is going to weigh less because there is not as much of it to weigh. If any of you are airplane pilots then you already know that the hotter the temperature the farther apart the air molecules are and the longer it takes to get off the ground and the slower you will fly. The same is true with vacuum. It will not hold down with as much force the hotter it is and the higher in elevation you are.
Remember this column of air is fluid so it will try to replace any vacuumed areas. This is very important. Air only weighs 15 lbs per square inch maximum. It cannot be made to weigh more with any size vacuum pump because the most you can vacuum out of anything is what’s there (15 lbs. max). If you have a 10" square you are trying to hold down the maximum you can hold down with is (10" x 10"=100 square inches) = 150 lbs of force. This is only with a pump that will pull 28-29 inches of vacuum. This can be accomplished with a positive displacement pump such as Busch, Becker, or others that can pull this kind of vacuum. Regenerative blowers on the other hand cannot pull this kind of vacuum but are more on the order of 10-12 inches which would make its hold down force per square inch more around 6 or 7 lbs per inch. If you were holding the same 10 inch sq piece of material it would be held down with more like 60 or 70 lbs.
Positive displacement pumps would probably be the ultimate but we have to remember that they do not put out the same volume as a regen blower. They put out much less per given HP because they are doing more work so you must find a balance. They also require more HP to get the job done. So if you use residential electricity you may find that you want to go with a 5-7.5 HP positive displacement pump such as Busch, Becker, etc and make sure you don't have leakages. I will mention some tips there later. If these pumps are too high new there are several companies that sell rebuilt ones with a guarantee. If you have plenty of electricity you may want a high volume regen blower such as Fuji, FPZ, etc of 20 or 25HP and when you have that kind of volume you don't have to be as bothered by leakage. On a 4 x 8 table I feel that 10 HP regen and below is a little bit marginal. Of course there are huge vacuums that will pull huge volume and high vacuum but most of us are trying to do CNC on a budget and could live off some peoples electric bills. Positive displacement pumps generally cost nearly 2-3 times the price of a regen blower.
I have a customer who has two of our machines. I was surprised to learn that he is running both tables with 7 1/2 HP rebuilt Busch pump that pulls 28-29 inches and has decent volume. He paid about 3300.00 for the pump. He is careful with his leakage.
Here’s how he does it. Our low-cost vac tables are MDF 1 1/2" thick with a grid system cut into the table with 4 zones on a 4 x 8. Each zone is 24" x 49" with a 2" PVC pipe running to each zone that has a PVC cut off/on for each zone. He mills a sheet of 3/4 LDF or MDF on one side. He then adhere this to his grid table being careful to only put the wood glue on the grid itself so as to not seal up the bottom of the sacrifice board. This keeps the sacrifice board from becoming sealed. Then he mills the top of the sacrifice board. He also then seals up the edge of the sacrifice board to not let air escape through the edges. When you cut your parts be careful to set your z so that it doesn't go unnecessarily deep into the sacrifice board. When the sacrifice board gets uneven and is affecting the vacuum you can then re-mill the board just enough to clean it up. When you get down to the grid again simply do the same with another board and start over. This works very well.
I have also found that if you have limited electricity available such as only 220 single phase and don't want a phase converted setup that a Dayton 10 HP blower working in reverse works fairly well because it has such a huge volume but uses a lot of electricity to start and you have to build a different vacuum box to handle the volume.
From contributor R:
That is a really good explanation of the vacuum system. The small parts are always an issue because the amount of hold down force is a direct function of the square inches of the part.
At a nominal 14.7psi for ambient air pressure, and accounting for leaks and general inefficiencies, I tell my clients to figure on about 10psi on their parts. So, the 4 x 4" parts mentioned above ought to have about 160 psi holding them. But, because they are small, and every cut through the sheet is essentially opening a new leak in your system, you need a better strategy.
I typically teach clients how to "onion skin" their parts. You make a primary cut, but not all the way through the part. Leave about 1/32" on the table. This skin keeps the parts from moving and keeps the vacuum seal intact.
A second cut is then made through the remaining skin. Since the cutter is only removing a tiny amount of material, there is very little side pressure trying to move the part. If you have a bunch of parts that are too small or complex for that technique, you can use material that is 1/32" thicker than needed and then flip the board over and run it through a wide belt sander. One or two passes and the parts fall off the conveyor separated from the web.
From contributor M:
Talk to the manufacturers. An oil ring pump will offer the least amount of maintenance. Vanes are a wear item, and you may have to replace them yearly, depending on your usage. I was quoted $1,300 per set a couple of years ago.
From contributor R:
Typically, I use the widebelt technique with parts being made from solid stock, not sheet goods, so getting the material 1/32 thicker is not an issue. Yes, tiny parts (like 32mm plugs) though a widebelt sander are likely to simply vanish into the dust collector. There are other strategies for such needs, like leaving a few thin tabs to keep the parts on the "web" until you remove them. I thought it was obvious that I was offering a few possible ideas, not one universal one for every situation. I have, in fact, used these and many other techniques over the years. And I'm quite certain that there are many creative woodworkers out there who are doing other things that might even work better.
The discussion was on the handling of small parts using vacuum. Since the original poster did not specify anything more than part size (4 x 4") I certainly was not recommending any particular way of dealing with hold down, just suggesting that there are a number of options.
Which path I would recommend is impossible to provide over a forum such as this. It would depend on too many other factors; how good is the vacuum, what other tools does the shop have that might help the process, and what is his labor cost? I have a client in Jamaica that would think nothing of having small parts hand cut from a larger sheet since his labor cost is virtually nothing.
From contributor R:
Again, all shops are different and will get different results. You are perfectly correct, I too would never run a 4 x 4 piece of anything through a widebelt sander. I will, and have, however, run larger pieces that have been cut mostly through, upside down, so that the sander removes the onionskin layer freeing the parts so that they drop free off the end of the belt.
As for justifying a CNC based on milling time vs. assembly time, or even material handling time, that too is dependent on the individual shop. Small, custom cabinet shops often are not under milling time constraints like high production shops. Assembly and finishing are often their bottlenecks. For larger companies that are replacing older machinery, the time saved from changing over from one setup to another can be their cost justification. And yes, for many companies, the rapid milling time is the cost justification.
Would you like to add information to this article?
Interested in writing or submitting an article?
Have a question about this article?
Have you reviewed the related Knowledge Base areas below?