Capscrew size calculation for pneumatic cylinder

[newby81]

Hi there,

I have been tasked with designing a pneumatic cylinder and was asked to do some calcs on what size bolts I should use to bolt the two endcaps of the cylinder to one another with the honed tube between them (Essentially I need to calculate if 4 off M10 capscrews will be good enough)

Unfortunately I never had the opportunity to complete any tertiary engineering qualification so my theoretical knowledge is abit limited. I am now in a situation where I feel like I am floundering abit and most of the guys at work are too busy to help or are not able to.

Please go easy on me with the explanations and a little patience please.

Some details are as follows:

The cylinder max force will be 9.2kN. There is no cushioning except for an o-ring in the face of the endcaps (The piston bumps into the oring which acts as an endstop bumper). We aim to run the cylinder fast (140 cycles per minute).

I have attached a rough model image to help convey the basic tie rod style design concept a little better.

Thanks

 

[MikeHalloran]

You will be better off in money, and time, by buying the cylinder you want from someone who makes custom cylinders for a living.

Many apparently simple products are not at all simple when studied in the detail necessary to design and build a successful one. Air cylinders are a good example.



Mike Halloran
Pembroke Pines, FL, USA

 

[desertfox]

I agree with Mike on this one, for a start the cylinder should be designed to some code and I would expect that code to detail the fastener calculation.
Further the cylinder should be pressure tested to ensure it doesn't fail in service and particularly a pneumatic cylinder as it could fail and go off like a hand grenade.

That said if you want estimate fastener loads,then workout the force from pressure times cylinder bore area and divide by four (for the number of bolts) this will yield the load per bolt.
Divide the force per bolt by the root area of the bolt and obtain the stress, which you can then compare with the allowed bolt stress.
What I have given you is a very rough approximation as a starting point but in reality it is much more complicated as you will have bending in the bolts due to the offset from the resultant force centre.

 

[CoryPad]

If you assume property class 10.9, then a single M10 screw has a minimum ultimate tensile strength of 60 kN, and a minimum proof force of 48 kN. It sounds like you are in a conservative regime, but you should do a complete analysis to be sure.

 

[drawoh]

newby81,

Are you are recent grad, or are you not an engineer? Is anyone knowledgable looking over your shoulder? The bolts are not the only thing you need to get right here.

--
JHG

 

[newby81]

Thank you for the information guys. @drawoh, unfortunately I am not a recent grad or qualified engineer. I have simply worked in the engineering design field for about 10 years now and have tried to learn what I could during this time.

There are other engineers in the department and the design is being assessed in design reviews before being manufactured for testing. Originally my design contained 4 off M12 capscrews, however, my engineering manager asked me to consider rather changing them to M10 but said I must do calculations to determine if M10 would be ok. I did not know how to do the calcs so thought I would seek help on this forum as the engineers at the office are often too busy to mentor me.

We would purchase the pneumatic cylinders and have done so prior to my employment, however, they were not up to the task of our machines and did not last so the company started making stronger cylinders themselves (And have been quite successful so far.)

Next week I will need to present the calculations to my manager and tell him if M10 will be ok.

Please guys like I said I am not qualified but I am keen to learn if there is anyone willing to help so any information is appreciated.

 

[SPDL310]

I hope you aren't sealing your cylinder with the same O-rings you are hitting 140 times a minute. Besides the static loading you may want to take the fatigue life of the bolts into account as you will see a peak stress each time the piston hits the end of the cylinder.

 

[CoryPad]

faq725-600
faq725-215
faq725-536
faq725-1461

 

[newby81]

Thanks again for the information. I will certainly check out the links you provided CoryPad.

@ SPDL310 The "bumper" seals on each end of the cylinder are both independent of the pressure chamber seals. I am certainly more concerned about the fatigue life of the bolts but just hope I manage to get enough information together and interpret it in the right way to calculate whether the bolt will hold up.

We have made triplex presses which can run at about 10.2kN and on those they have used M10's successfully but on this press the piston is heavier so I suspect there will be a higher impact load on each cycle. I think the M10's will be ok but my manager likes numbers as do I provided they are right that is.

 

[Tmoose]

Then there is the installation torque and resulting clamp force that must be high enough to keep the tie bolts from stretching and unloading the seals ( o-rings?) between the end caps and the cylinder, and also to reduce the risk of fastener fatigue.

 

[desertfox]

Hi newby81
If you read the second part of my first post I've outlined the starting point for your calculations.

If your maximum load is 9.2KN then divide this by the four bolts which 2.3KN per bolt, now a grade 10.9 M10 has a proof load of 48.1kN according to this site

http://www.roymech.co.uk/Useful_Tables/Screws/Screw_loads.htm.

Now what we don't know is the cylinder material yield strength because if you tighten the cap screw to tight the cylinder material will yield and bolt preload will be lost.

 

[tbuelna]

Those cap screws will need to have sufficient preload to prevent fluid leakage under the applied loads at each end fitting from fluid pressure, dynamic forces from travel stops, and moments on the end attachments.

 

[newby81]

@desertfox.

I finally managed to get some numbers down on paper. Please could you have a look at the attachment and let me know if I have understood your guidance correctly. Could you clarify the following for me please?

Firstly I need to compare the effective force per capscrew to the proof load for the capscrew. (ie. My effective force per capscrew is 2.3kN and the proof load for a grade 12.9 M10 is 56.3kN, therefore this comparison suggests that the M10 is well within its limits for this application.)

I then calculated the effective stress per capscrew to be 45.19MPa. I compared this value to the yield strength of the endcap materials (three to choose from)

Mild Steel - 300 MPa
Aluminum 6082-T6 - 250 MPa
Aluminum 6083-T6 - 214 MPa

All of the materials have a yield strength far higher than the effective stress per capscrew and I would therefore also assume that this further supports indications that M10 capscrews are well within their limits for this application with all the endcap material options.

Are my assumptions above correct and have I interpreted your guidance correctly?

Thanks

 

[zeusfaber]

Is the actuator rod attached to the load, or does it just press on it? If the first, you will need to consider the forces involved in bringing the load (as well as the piston) to an abrupt halt at the end of the stroke.

A.

 

[desertfox]

Hi newby81

Yes your on the right lines with the calculations you've presented.
Now if the cylinder load is a maximum of 9.2kN then the the combined preload of the four screws needs to be larger than this, so let's say the preload for each screw is 4.6kN, that should ensure that the end cap doesn't move under load.
Now to determine whether the clamped parts yield you need to calculate the area under the screw head which is in contact with the end cap and divide that by the 4.6kN preload, if this exceeds the yield stress for that clamped material then you will need to add more screws so that the preload can be reduced to avoid yielding the clamped components.
Technically you should be following some design code for the design of the cylinder because you might well have to consider bending of the end cap and/or compression of seals/gasket.