We are designing robot that will pick and place sheets of paper. We have found an issue that it is hard to suck only first sheet of paper of paper pack. Issue occurs because paper sheet passes vacuum through it and vacuum sucks additional (2 - 3) sheets of paper.

How to resolve this? Maybe we are using too small suction cups or incorrect design?

Current pneumatic system contains following: 1) Vacuum pump that produces 120 kpa vacuum 2) SGPN 20 FPM-65 suction cups 20 mm in diameter for paper handling 3) Need to hold standard office paper A4 80 gsm (210 x 297 mm) 4) Distance between suction cups is 100 x 125 mm 5) Paper pack could contain at least 30 sheets of paper.

Current experiments results. 1) 1 suction cup easely holds 5 sheets of paper because vacuum is flowing through paper 2) if pressure is decreased to minimum its stops sucking or sucks 3 sheets

Solution which we are trying to get is shown on this video: https://www.youtube.com/watch?v=FT7NWbXyM7Q

Similar issue that we are getting is shown on this video: https://www.youtube.com/watch?v=LKBDzkm3moo

Maybe somebody have other ideas how to make it possible to suck only first sheet of paper without mechanical holding of it?

Suction cups example

Paper suction cup

Paper suction

  • $\begingroup$ Need some more info here...weight of paper (as in g/m^2) how many suction cups, their distribution and the intended vacuum at said cups. There could be many issues...static comes to mind as a likely culprit... $\endgroup$
    – morbo
    Jul 29 '19 at 21:57
  • $\begingroup$ Paper weight 80 gsm. We are using 4 suction cups. Distance between cups is 100 x 100 mm. For suction we are using vacuum pump with max pressure 120 kpa. Pressure is more than enough for that purpose. $\endgroup$
    – Andris
    Jul 29 '19 at 22:10
  • $\begingroup$ And the intended vacuum at each cup...in bar? $\endgroup$
    – morbo
    Jul 29 '19 at 22:11
  • $\begingroup$ Looks that is is 120 / 4 = 30 kpa for each cup that makes pressure on 20 mm circular area. $\endgroup$
    – Andris
    Jul 29 '19 at 22:13
  • 1
    $\begingroup$ how do laser printers pick a single sheet from a stack of paper? $\endgroup$
    – jsotola
    Jul 30 '19 at 0:53

As I suggested, there can be several issues going wrong.

One of them is static, this can definitely cause issues. Though it's hard to know if that's specifically yours or not.

The other is your design is overkill.

I assumed you're trying to pick up a sheet of A4, 80gsm paper.

Do some basic mathemagics:

$$\text{sheet}=\frac{(0.08\text{kg}) \times\left(62370\text{mm}^2\right)}{1\text{m}^2} == 0.0049896\text{kg}$$

Add some gravity,

$$\text{sheet} \times\left(9.81\text{m}/\text{s}^2\right) == 0.048948\text{N}$$

Now calculate the force each cup is providing..with giving comment infos...

$$(30\text{kPa})\times \left(100 \pi \text{mm}^2\right)== 9.42478\text{N}$$

My first suggestion is drastically reducing your vacuum...30kpa at each cup is just way too much. After doing that, one could also consider other things..reducing cup size, spreading out your cups to the corners...dealing with possible static, etc etc.

  • $\begingroup$ morbo, if we will use larger size of suction cups that will make suction force bigger? $\endgroup$
    – Andris
    Jul 30 '19 at 20:17
  • $\begingroup$ If you can maintain the vacuum (namely 30kPa) with a larger cup/surface area, yes. $\endgroup$
    – morbo
    Jul 30 '19 at 20:44

I think the primary problem you are seeing here is that you are trying to pick up ordinary paper, which is relatively porous. When I was working with suction cup systems we were dealing primarily with labels, so non porous, waxed backing paper. Our page turning systems used rubberised rollers rather than suction cups, even though we used pneumatics elsewhere in the page turning mechanism.

As the two videos show, with the waxed paper only one sheet at once is usually picked up, whereas with the more porous tissue, several are usually picked up. In fact the vacuum is so strong that with only a single tissue, the demonstrator was having trouble separating the suction cup from the table below.

I agree with morbo that your system has way more suction than it needs, but I think it's probably easier to find the right balance experimentally than to try to work out from first principals, with all of the variables that aren't accounted for by the math. I think you need to get a stack of paper and your actuator (and possibly some alternatives) and find the limits of your system.

You need to be able to control the vacuum

Without being able to control the vacuum you are going to have a hard time controlling the behaviour of the system.

As in the videos, you need to be able to see what the vacuum pressure is when picking up and holding the paper, so you need a relatively accurate vacuum gauge.

Depending on your characterisation of the system, you may only need variable vacuum control while performing that characterisation, but it may be that you need to vary the vacuum dynamically during different phases of your pick & place operation. Rather than just on/off you may need low vacuum to pick and lift a single sheet, followed by a switch to high vacuum befor a fast move to the place location without dropping the paper.

How a variable vacuum is achieved is beyond the scope of this answer, but may involve a variable vacuum source, a variable regulator, a bypass or some king of pulsed source or bypass. The specifics are beyond my expertise. What I do know is that you need to characterise the behaviour of your system.

Characterise the behaviour of your suction cups

Starting from a pressure too low to pick up any paper steadily increase the vacuum until it picks up one sheet. Then keep increasing the vacuum until it picks up two sheets. The vacuum at the mid point between the two should give you a good quick compromise between holding a single sheet securely and the risk of picking up a second sheet. Depending on the reliability you need though (99%, 99.9%, 5 nines) you may need to characterise your system more thoroughly.

Repeat a pick-up attempt at each pressure enough times to build up confidence in how much variation there is in results. If vacuum $x$ picks up 1 sheet 99% of the time and never picks up 2 sheets, while vacuum $y$ picks up 1 sheet every time, and 2 sheets 1% of the time you can surmise that you have a well characterised system and can probably rely on the midpoint to pretty reliably only pick up 1 sheet.

If vacuum $x$ picks up 1 sheet 80% of the time and 2 sheets 20% of the time, while vacuum $y$ picks up 1 sheet 20% of the time, and 2 sheets 80% of the time, then you may have to consider whether this is the right solution for you.

Check it behaves the same at the extremes

It may also be worth checking if behaviour is different when the stack of paper is full compared to when it is empty. If not, then you may have to chose a different compromise point to maximise the reliability over the whole paper refill lifetime. Alternatively, you may have to calculate vacuum pressures at the pick location depending on the number of sheets in the stack (if you can't mechanically compensate for the stack height, as printers do).

Finally, you may want to try running the same characterisation with 70gsm and 90gsm paper as well as at 80gsm. While there is little variation in commercially bought paper, you can never rule out customers choosing to buy cheaper thinner paper, or decide they need to use heavier paper stock for some jobs.

In my experience it's better to know about likely problems before they occur than have to deal with them in the middle of the night.

Frantic Customer: We filled the machine & it started picking up 2 sheets at once!
Bleery eyed Engineer: What did you change?
Apologetic customer: Oh, this job needed 70gsm paper
Resigned engineer: Ok, go into the .ini file and reduce the vacuum value by 20% then on Monday put in a change request to implement 70gsm paper support in the GUI


Reliability only comes from testing, and testing again after every adjustment you make. Data gives you the ability to reason about the behaviour of your system, predict what changes will improve or impact on reliability and test hypotheses until you get what behaviour you need, or decide you need to a different solution.


I have in mind the following:

  1. When placing the stack of sheets, make sure that they are not bound at the edges, due to the cutting. Just "wave" the stack of sheets (the edges are enough) to separate them.

  2. Move the top-most sheet horizontally (sliding movement) using some rubber rolls. This system is used in pretty much all printers.

  3. Pick the separated sheet of paper. Or keep sliding it until it reaches the next machine.


Sounds like some kind of stiction, between individual paper sheets.

How about:

  1. Suck three or five sheets.
  2. Do some secondary manipulation to remove the additional sheets. Like lateral motion, twisting etc. Think about how you get rid of extra sheets when turning a book page.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.