Back pressure on a pipe by squeezing it shut

[beds87]

Hi all,

I'm new to the forum and was after a bit of guidance.

At work we are looking into purchasing a new dosing unit, however I'm unsure if the design that has been drawn up is going to work.

The units design revolves around the principle that a flexible tube is compressed closed by rollers actuating against the tube.

The rollers are set below the tube offset below the tube and when actuated close parallel to each other completely closing the tube.

My question is would those rollers coming together create a backwards pressure? If so how would I calculate it?

Some of the product will be left below the rollers in the flexible tubing with no valve stopping it from dripping, so if there is backwards pressure, for the product not to drip the pressure must be greater than the specific gravity of the product? Is that correct?

If a drawing is needed to further illustrate the question I can drawn something up.

Thanks

 

[BvU]

Hello Beds,

Not sure I understand your scenario. I imagine the kind of hoe pumps the guys in my village produce too. Like these.
At what moment in the cycle do you worry about back pressure ?

 

[beds87]

I quite possibly didn't explain myself very well so I have drawn it up.

Would the roller rotating upwards and closing the pipe work create a backwards pressure to hold the product below the rollers within the nozzle with enough force to stop that product dripping out? There is no mechanical valve to stop the product dropping it's just a straight through hole. The nozzle isn't submerged in product so Air can travel freely into the pipework when the rollers actuate to clea the pipe.

%5BIMG%5Dhttp://i64.tinypic.com/2jeuut4.jpg%5B/IMG%5D

%5BIMG%5Dhttp://i66.tinypic.com/1429281.jpg%5B/IMG%5D

I hope these images work. If not what's the protocol for uploads?

Thanks

 

[BvU]

I hope these images work

Not for me they don't. I get an invitation to join in order to upload things, neither of which I want to do.
For uploading pictures there is a button on the lower right.

 

[beds87]

Hopefully this works.

 

[BvU]

Yep, it works (the uploading). Back- and forward presssure seem unavoidable when the roller comes down, even when it's moving to the left. But I 'm not sure I have your question the right way: are you worried about dripping on the right (supply side -- in my view connected to a reservoir or something) or about dripping on the delivery side (the left if I am not mistaken) ?

This delivery side should be less of a problem, since when the rollers separate again, the same phenomoenon occurs in the other direction: a small underpressure when the tube reassumes its normal volume.

 

[beds87]

For some reason it has flipped the photo while uploading the left side is the sump side, this is not left to right but actually top to bottom.

Left side is where the product is delivered to via a servo drive so an exact cc is delivered. The PLC then triggers the rollers to actuate and close.

What I'm trying to calculate is when the rollers close will the pressure caused by the pipe being blocked stop the product which is already below the rollers and in the nozzle from dripping?

Basically I need to calculate the back pressure from the rollers closing then use that against the specific gravity of the product to find out what size opening that should be used on the nozzle at the end of the pipe?

Does that sound possible?

Thanks

 

[BvU]

Ok, so feed is from left in the picture as it appears on my screen now. Dosage stops and probably blocks backflow to the left. Any volume reduction of the tube necessarily pushes product to the right. Gravity is one thing, viscosity the other.

What I don't understand is the logic: if the dosage pump stops, why bother to use these rollers ? Afraid air is leaking in ? If so, what's the fundamental change between the two pictures that would prevent such a mishap ?

@Chestermiller : have you got any good ideas on this, Chet ?

 

[beds87]

The reason for the rollers is that there are four of these dosing units fed from a manifold which is higher than the dosing units. Although there is no applied pressure from the pump on the product it seemed to drip out more freely due to the amount of product in the pipeline going back to the manifold.

The initial idea of the rollers was based around the principle of holding water in a straw when you keep your thumb on the top end.

We have done a few simulation trials and it's a bit hit and miss. If the nozzle size is too small it sprays the product out all over the place but the product doesn't drip. If the nozzle size is too large it drips out.

The issue is we run multiple products through this and the specific gravity can change between batches of the same product let alone product changes.

 

[Chestermiller]

If the tubing is close to being perfectly vertical, then surface tension should help to support a column of liquid above the exit, without drops forming at the exit. The surface tension force at the verge of a drop forming is $\pi d \sigma$, where d is the nozzle diameter and $\sigma$ is the surface tension of the fluid. For water, the surface tension is 70 dynes/cm.

 

[beds87]

So to calculate the nozzle diameter. I'd first need to calculate the surface tension of the liquid using;
S = (ρhga/2)

Then re arrange your formula above to give d

Is this correct?

 

[Chestermiller]

So to calculate the nozzle diameter. I'd first need to calculate the surface tension of the liquid using;
S = (ρhga/2)

Then re arrange your formula above to give d

Is this correct?

Yes. This is a rough approximation.

 

[beds87]

To calculate this for each batch someone would need to place a tube into a known level of liquid and determine how higher past the original level the product had reached.

Is there a more simple solution?

 

[Chestermiller]

You don't need to do the experiment in the exact tubing setup that you are using. You can set up a test rig, even without the roller. Do you know the properties of the fluid (density and surface tension) and the diameter of the nozzle? This might help to determine a critical value of surface tension divided by (density x diameter). What I'm saying is doing off-line experiments.

 

[beds87]

You have now lost me.

All that is currently tested on each batch of product is the weight and specific gravity. Is there a way to calculate surface tension using the specific gravity?

Any further tests I would need to implement.

I know the weight so could calculate density.

I was then going to set up a test rig and place a known radius capillary tube into a product and measure the excess height over the product surface within the tube. Then calculate the surface tension using:

S = (ρhga/2)

An then transpose the equation you supplied to find d the diameter of the nozzle which will have enough surface tension for the product not to drip.

Are you suggesting there is a simpler method?

This forum has been very informative, thanks for all the support in your replys it's much appreciated.

 

[Chestermiller]

You have now lost me.

All that is currently tested on each batch of product is the weight and specific gravity. Is there a way to calculate surface tension using the specific gravity?

Any further tests I would need to implement.

I know the weight so could calculate density.

I was then going to set up a test rig and place a known radius capillary tube into a product and measure the excess height over the product surface within the tube. Then calculate the surface tension using:

S = (ρhga/2)

An then transpose the equation you supplied to find d the diameter of the nozzle which will have enough surface tension for the product not to drip.

Are you suggesting there is a simpler method?

This forum has been very informative, thanks for all the support in your replys it's much appreciated.

There are methods and equipment for measuring the surface tension directly. But, what I am suggesting is that you set up a separate test rig to just test each solution. It could be a length of tubing with a nozzle attached. You could have atmospheric pressure above the liquid in the tubing (the pressure at the rollers in the actual equipment would likely be sub-atmospheric). This equipment would determine the maximum length of column to avoid leaking.

 

[beds87]

I understand what you are saying now.

So I could:

Before each batch is pumped out I could use a tube the same length and diameter as the dosing unit tube, place a nozzle on, dip it into the product and block the top end and drop nozzle sizes until it doesn't drip.

 

[Chestermiller]

I understand what you are saying now.

So I could:

Before each batch is pumped out I could use a tube the same length and diameter as the dosing unit tube, place a nozzle on, dip it into the product and block the top end and drop nozzle sizes until it doesn't drip.

Not exactly. Pour liquid into the top of the open tube until it starts to drip.

 

[CWatters]

Why use rollers to pinch the pipe rather than a regular lever valve?

 

[CWatters]

How about using a valve arrangement that shuts off the fluid and expels any remaining fluid below the valve using compressed air or water?

 

[beds87]

The issue is we are dosing 200 bottles of product per min of varying specific gravity and weights. We have tried a wide array of mechanical valves yet nothing can keep up or doesn't drastically wear over a short period.

When the nozzle size is correct to the product there are no leaks. It's just establishing the correct nozzle size to the product batch as there are slight variations between batches causing the same nozzle to drip or even leak.

 

[beds87]

Those 200 bottles per min is shared between 4 nozzles so each nozzle operates 50 times a min

 

[CWatters]

So the rollers will be compressing the pipe 200 times a min? Will the pipes stand up to that?

 

[beds87]

The pipework does need to be changed every few days but the cost and time of swapping a pipe out against a PLC controlled festo solenoid valve. The pipework can also be changed when each product is swapped out and then cleaned out when the nozzle is changed.