Force Required to Open Door

[alister1]

Hi All,

I am currently looking to specify a hydraulic cylinder and power unit to open and close a horizontal pressure vessel door.

The door weighs 566 kgs and is mounted via a hinge pin that is sat on a thrust bearing (SKF 51202).

The cylinder will be mounted 175mm from the hinge pin at the bottom of the door.

Can anyone suggest the calculations that will be required to solve this? i have attached a basic sketch and the datasheet for the bearing can be found here:

http://www.skf.com/uk/products/bear...single-direction/index.html?designation=51202

Thank you in advance.

 

[hydtools]

How fast?
The door swings horizontal and the hinge is vertical?
Can you draw a free body diagram?

Ted

 

[alister1]

Speed to open to 95° would need to be approximately 20 seconds

Yes the hinge is vertical with the door opening horizontally.

The diagram would be similar to the image below with the force been applied 175mm away from the hinge.

Hope this helps.

Thanks.

 

[hydtools]

I think you can look at the problem made up of two stages. The first stage is a constant acceleration stage. The second stage is a constant rotation speed stage.
First stage: impulse = momentum M*t = I* w; M = moment = force * r*cos(theta), t = time of constant acceleration, I = mass moment of inertia of the door, w = rotational speed, theta = angle of door rotation during acceleration

Second stage: constant speed theta2 = w*t2; theta2 is rotation angle during constant speed, w = rotational speed, t2 = time of constant speed rotation

t + t2 = 20 secs
theta + theta2 = 95(pi/180) radians
M = F*r*cos(theta) For a small angle during acceleration M = F*r

F = (I*(95 - theta)*pi)/(180*r*t*(20-t)) Select values for acceleration time t and acceleration angle theta(in degrees) to calculate force F. Repeat with other values of acceleration angles and acceleration times as required to find sensible hardware values.

Ted

 

[kiwitom235]

Hello. Basically you would need to size your cylinder and pump so that the time it takes for cylinder to fully extend is 20 seconds.

I image force required to push the door is not overly high as it is rotating about a vertical axis.

P= F x v. So you can work out your motor power rewuired from force and velocity. taking about 80% efficiency. Will need to use the power equation for angular velocity though.

You then need to specify a cylinder size and pump size from there...

 

[TimSchrader2]

Hello

The 20 sec extend speed will govern the size of your pump and motor. Most hydrl pumps have a press relief so you can set the pressure where it needs to be. So the force req can be somewhat aprox, in this case. If $$ are a big deal you can get small pumps that put out only 500 to 200psi. Some of these may not be adjustable with pressure.

Estimating the force req is still needed to get some idea of your min bore side. Per above methods. The retracting will require more pressure. Then you can see how close your calc is to actual press after you test it.

I doubt if the cyld will be at 90deg to the door, so you lose a bit of effective turning force.

 

[capliard]

What I see missing is the pressure difference between side. It's a pressure vessel, there has to be a pressure differential.

In mining, we often use a small length but large diameter strong cylinder in conjuction with a longer but leaner weak cylinder.

The small one is there to help ''break the seal''. Once the pressure diffence is gone by opening the door, the longer, leaner one takes over.

The other option is to break the seal by having a parallel opening that opens before the larger door. Usually a slide door.

The door weigh is negligible as it does not work against the cylinder. Once the seal is broken, the door will open as fast as your cylinder can slide.

I've only seen systems with pneumatic cylinders. We were working with 80-100 Psi, so we increase cylinder diameter to add force.

Keep the hinges lubed up and you'll be alright.

If pressure difference is large between two sides, consider a double egress door, which is a double pannel door with a pannel openning North and a openning South. One end of cylinder fixed to each door. It works well too.






Ingenieur Minier. QuTbec, Canada.

 

[chicopee]

There is one more time element probably missing and that is the decelaration time when the door is to stop when reaching its full open position.

 

[hydtools]

Chicopee, yes there could be deceleration time, or just let it bang against a stop.

Ted

 

[PNachtwey]

How precise does this need to be? Why 20 seconds?
All those equations are not going to be much help without a motion controller and if you get a decent motion controller you don't need to do the math. You simply build a cam table that does a third order interpolation between the linear position of the door in degrees to the extension of cylinder in millimeters.
When the door is at 0 degrees, enter 0 and the current position of the cylinder. Then job the door to about 10 degrees. Use a level to measure the angle. Then enter the cylinder position for 10 degrees. Repeat until 95 degrees. Now the user can command the door to move in units of degrees, degrees/s, degrees/s^2.
The motion controller will take care of using the door position in degrees and indexing into the cam table or curve table to do a cubic interpolation for position in mm, velocity in mm/s and acceleration in mm/s^2. There are simple equations I use for computing speeds and acceleratoins
speed = (3/2)*(distance/time) = 7.125 degress/second
acceleration = (9/2)*(distance/time^2) = 1.096 degrees/second^2
This assume one third of the time is accelerating, one third at constant velocity and one third of the time decelerating. This will be a very smooth move.

We have used this techique on many years.
If you want we can calculate this precisely and take the derivative to get the velocity as a function angle and the second derivative to get the angular acceleration. The controller is smart enough to apply the chain rule.
It is also possible to change the controller gains as a function of angle every millisecond or faster.





Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[hydtools]

Too bad we have not heard from the OP.

Ted