Controlling two valves with different flow-rates to achieve mixing

  • #1

Main Question or Discussion Point

Guys,
I hope you guys are keeping well.

I have been working on developing a Compact Dispensing Station for my CNC Machine & Parts Washer where coolant is mixed with water in different concentrations & volumes as per the requirement which will vary on a day-to-day basis (not the concentration, but the volume).

My design consists of two lines with valves & Flow meters, one for coolant (pumped from a barrel/drum) and the other for water (coming directly from the plant's water supply pipes). Concentration of coolant required is in the range of 1-3% for top-up of sump & liquid volumes(mixed) required will be different everyday, as per the losses in the CNC sump. But normal range I am expecting would be around 50 Litres +- 10/20%

My system consists of a Direct Acting Solenoid Valve followed by a Flow-meter which detects liquid flow-rate & sends signal to the PLC to shut-off the valve when desired flow is met. This is in both the lines (coolant & water).

Now since a 2% coolant for 50Litres would mean approx. 1 Litre of coolant in 49 Litres of water, how do I control the flow of both the liquids so that I achieve homogeneous mixture through-out. I cannot run both the liquids together cause then how would I meter the flow of coolant to match the flow of water?

I do not have space for another tank where I can drop both these liquids and achieve mixing.

I am using this as reference - (Skip to 4:10 in the video).

Any help will be appreciated.
Thanks,
KV
 

Answers and Replies

  • #2
245
172
I don't think that you can do (very well) what you want with the hardware that you describe. You could (theoretically):

Regulate pressure of both fluids
Set the flows to known/constant values with manual metering valves
Turn on the water
'pulse' the coolant ON/OFF based on computed ratio for desired mixture
dispense the mixture through a static mixer large enough to make the 'pulses' negligible

Getting the relative flows adjusted so that you can cover your expected dynamic range of concentrations is the critical thing. This will never be as precise (for several reasons) as:

Buy a dispensing pump for the coolant. You can measure water flow and 'slave' the dispensing pump to maintain the desired concentration. No pulses. A very small static mixer may (or may not) be required.
 
  • Like
Likes russ_watters
  • #3
Thanks for replying.

Could I do it in this fashion? For eg. If I have to top-uip 50 Litres @2% i.e 1 Litre of coolant with 49 Litres of water.

This data shall be fed to the PLC prior to dispensing. So the PLC will automate the opening & closing timings by dividing each flow-rate into equal 'pulses'.

For every 1 Litre of water the coolant pump will dispense 1/49 = 20ml and then the valve would switch off. The moment the flowsensor on the water line senses 1 Litre has passed, the pump will kick back again and pump another 20ml and so on.

Will this work? Which pump would be a good fit for this application?


You can measure water flow and 'slave' the dispensing pump to maintain the desired concentration
Could you explain this please?
 
  • #4
jrmichler
Mentor
1,128
1,130
You don't state the size of the CNC coolant sump, or the tolerance for coolant concentration, so take this suggestion accordingly:

Assume the coolant sump is 500 liters, and you want to add 50 liters of 98% water / 2% coolant. Also assume that you have 98% water / 2% coolant in the sump. If you were to add 49 liters of water, followed by 1 liter of coolant, and do this while the machine coolant pump is running, then the coolant concentration would start at 2.0%, drop to 1.8%, and then rise to 2.0%. Or you could add the coolant when half the water is added, so the concentration would start at 2.0%, drop to 1.9%, rise to 2.1%, then drop to 2.0%. This assumes that the coolant is pumped in a roughly the same rate as the water.

It is usually a bad idea to control small amounts by running the pump for a very short time. Start / stop losses and errors can cause huge total errors. Your coolant addition pump should be small enough that it runs for at least 10 seconds each time, and a full minute would be much better. You might consider a peristaltic pump for coolant addition. A peristaltic pump is a positive displacement pump, so the total flow is proportional to pump speed times flow rate, making a flow meter unnecessary.

Another suggestion:
A small positive displacement coolant addition pump with a variable speed drive could set the coolant addition rate at any desired percentage of the water addition rate. Then you would run the water addition simultaneously with the coolant addition, and there would be no changes in concentration. There are also metering pumps that allow you to set any desired flow rate, with the flow rate adjustable manually or through a PLC.
 
  • Like
Likes kunalvanjare
  • #5
You don't state the size of the CNC coolant sump, or the tolerance for coolant concentration, so take this suggestion accordingly:

Assume the coolant sump is 500 liters, and you want to add 50 liters of 98% water / 2% coolant. Also assume that you have 98% water / 2% coolant in the sump. If you were to add 49 liters of water, followed by 1 liter of coolant, and do this while the machine coolant pump is running, then the coolant concentration would start at 2.0%, drop to 1.8%, and then rise to 2.0%. Or you could add the coolant when half the water is added, so the concentration would start at 2.0%, drop to 1.9%, rise to 2.1%, then drop to 2.0%. This assumes that the coolant is pumped in a roughly the same rate as the water.

It is usually a bad idea to control small amounts by running the pump for a very short time. Start / stop losses and errors can cause huge total errors. Your coolant addition pump should be small enough that it runs for at least 10 seconds each time, and a full minute would be much better. You might consider a peristaltic pump for coolant addition. A peristaltic pump is a positive displacement pump, so the total flow is proportional to pump speed times flow rate, making a flow meter unnecessary.

Another suggestion:
A small positive displacement coolant addition pump with a variable speed drive could set the coolant addition rate at any desired percentage of the water addition rate. Then you would run the water addition simultaneously with the coolant addition, and there would be no changes in concentration. There are also metering pumps that allow you to set any desired flow rate, with the flow rate adjustable manually or through a PLC.
Thanks for this.

Just one query. Industry experts recommend addition of the top-up liquid in a mixed emulsified condition. Adding water & coolant alternatively might mess with the emulsion in the sump.

Also, this will be more substantial in the case of using my machine for not just top-up, but re-filling the entire tank. Imagine filling 400 Litres @5% i.e 20 Litres of coolant with 380 Litres of water. Do you think alternatively adding water & coolant will result in a homogeneous mixture in the tank?

Thanks for the suggestion on the Peristaltic pumps. Are these used for Coolants though? Also I will want to mount the pump on my system, with a suction hose that will go into the Coolant Drum, length of this hose could be as long as 3 metres.

Also, I have looked around for Peristaltic pumps, but mostly what I am seeing are small pumps that look suitable for lab scale use. Any known brands you can suggest?
 
Last edited:
  • #6
jrmichler
Mentor
1,128
1,130
Also, I have looked around for Peristaltic pumps, but mostly what I am seeing are small pumps that look suitable for lab scale use. Any known brands you can suggest?
Try searching using search term industrial peristaltic pump. One of the first hits has sizes to 475 GPM and suction lift to 9 meters. Back when I worked in a paper mill, I sized and purchased over 100 pumps, but do not have experience with peristaltic pumps.

There are other types of positive displacement pumps with suction lift capability that might meet your needs. I suggest contacting some pump suppliers, and discussing your needs with an application engineer. The time to contact an application engineer is after you determine exactly what fluid you are pumping, flow rate or range of flow rates, pressure, and suction lift. Those people do nothing but find the best pump for an application.
 
  • Like
Likes kunalvanjare
  • #7
Just one query. Industry experts recommend addition of the top-up liquid in a mixed emulsified condition. Adding water & coolant alternatively might mess with the emulsion in the sump.

Also, this will be more substantial in the case of using my machine for not just top-up, but re-filling the entire tank. Imagine filling 400 Litres @5% i.e 20 Litres of coolant with 380 Litres of water. Do you think alternatively adding water & coolant will result in a homogeneous mixture in the tank?
Could you shed some light on this, please.
 
  • #8
Baluncore
Science Advisor
2019 Award
7,732
2,684
A peristaltic pump or metering pump is needed for introducing the coolant. A peristaltic pump could also be used to meter the water.

If you do not have space for a second tank, then you cannot premix the coolant. The ratio of coolant to water may change when topping up the tank to maintain concentration. That makes a premix difficult. Do you measure the density to estimate the concentration?

Adding water & coolant alternatively might mess with the emulsion in the sump.
I would meter the coolant into the main tank, then add water. For the first fill I would alternate coolant and water while gradually increasing the volumes. Do you have a mechanical stirrer or a circulation pump in the main tank that can mix the coolant as the volume is increased?
 
  • #9
Do you have a mechanical stirrer or a circulation pump in the main tank that can mix the coolant as the volume is increased?
There is no stirrer, but a circulation pump is available in the machine that transfers coolant from the sump to the nozzles in the machine. I am aware mixing of coolant would happen by running a few circulation cycles before using the machine for production, but that would also lead to some losses. I would ideally want the coolant to be mixed with water before adding into the sump.
A peristaltic pump could also be used to meter the water.
Any reason why? The water will be fed through the plant's water supply line. There is no need(or space) for another tank.
Do you measure the density to estimate the concentration?
We have a refractometer.
 
  • #10
Baluncore
Science Advisor
2019 Award
7,732
2,684
Any reason why?
A peristaltic pump is actually a compact flow meter that you control. If the input comes from the plant's water supply line, you have full control of the flow rate and volume.

I would ideally want the coolant to be mixed with water before adding into the sump.
Ideally; perfection is the enemy of progress.
But you have no extra tank and no mixer, so you must mix in the combined outlet from the two inlets. The coolant ratio can be mixed in real time by using significantly different sized pumps and tube.

If you provided a cross connection 3-port solenoid valve between the circulation pump outlet and the sump you could mix with the existing pump, in the existing sump.
 
  • #12
Svein
Science Advisor
Insights Author
2,080
669
'pulse' the coolant ON/OFF based on computed ratio for desired mixture
Be very careful when "pulsing" solenoid valves. They are not designed for such operation, and you might introduce pressure waves in the tubes.

Anecdote (yes, another one, stemming from 40 years in the R&D profession): Once we were working on a huge robot (actually a tunnel-bore rig, but anyhow). The guys working on the hydraulics had installed proportional control valves, but they had a tendency to stick at very low control voltages. So one of the electronic guys came up with the idea of "shaking" the valve when the control voltage was low. I saw the aftermath: One electronics guy, very wet with hydraulic fluid uttered the eternal truism: "When an electric line breaks, the current stops flowing. When a hydraulic line breaks, the current starts flowing."
 
  • #13
245
172
I'm not sure I agree that pulsing (open, close) a solenoid valve is a problem. By 'Pulsing,' I only meant relatively short open intervals (not attempting PWM).

Your electronics guys were just ahead of their time. 'Shaking' proportional valves by means of a (relatively) high frequency dither on the base control signal is exactly how the valves in most transmissions work - that keeps them in the dynamic (not static) friction regime.
 
  • Informative
Likes anorlunda
  • #15
https://www.dosatronusa.com/
-
I'm thinking this is being made way more difficult than necessary.
This does not account for changes or fluctuations in concentration as it "doses" the same proportion everytime. Which is not practical for my application.

Employing a person to change the setting on the Dosatron defeats the purpose. I have also looked at Venturi Mixers & Coolant Proportioners. But they all come with certain degree of limitations, plus availability is an issue here in India.
 
  • #16
Once you get the metering figured out, I think a Static Mixer is what you are after. Here is one example:
https://www.statiflo.com/statiflo-s...component-for-disinfection-of-drinking-water/
Yes a Static Mixer is being considered for mixing. But again, the pressing issue is over how to introduce the two liquids into the mixer.

Please consider the following :-
Case A : You match the flow-rate of the two liquids so that both liquids flow through a Static Mixer at the same time. For eg. if you're adding 1 Ltr of coolant with 99 Ltrs of water, you control the flow-rate of water to match that of the coolant so they both flow together throughout the entire duration of the dosage, and finally mix and end up in the tank for a 1% concentration mixture.

Case B : You let the two liquids flow together at their own respective flowrates. The 1 Ltr of coolant passes through the mixer with water in 5 seconds and then the coolant pump shuts off, while the water flow continues till 99 Litres pass through. Would I still end up with a 1% concentration and will the mixing be homogeneous?

Some insight on this would be great!
 
  • #17
Another query, will the pressure difference between the two flows i.e coolant (through pump) and water (under gravity) create a problem with mixing?
 
  • #18
A peristaltic pump is actually a compact flow meter that you control. If the input comes from the plant's water supply line, you have full control of the flow rate and volume.
My plan is to only use a pump for transferring coolant from the drum into my system. Water will be introduced under gravity. How would a Peristaltic pump be better over a Gear pump for example?
 
  • #19
Baluncore
Science Advisor
2019 Award
7,732
2,684
I have been working on developing a Compact Dispensing Station for my CNC Machine & Parts Washer where coolant is mixed with water in different concentrations & volumes as per the requirement which will vary on a day-to-day basis (not the concentration, but the volume).
I get confused by the terminology in this thread.

A “cutting fluid” is usually a liquid emulsion made from a soluble oil mixed with water. The soluble oil is the cutting compound and lubricant, while the water cools the cutter.

The sump of a CNC machine collects the used cutting fluid for cooling and re-circulation.

The concept of a “parts washer” is inappropriate, because it is the cutting tool that is being irrigated.
 
  • #20
I get confused by the terminology in this thread.

A “cutting fluid” is usually a liquid emulsion made from a soluble oil mixed with water. The soluble oil is the cutting compound and lubricant, while the water cools the cutter.

The sump of a CNC machine collects the used cutting fluid for cooling and re-circulation.

The concept of a “parts washer” is inappropriate, because it is the cutting tool that is being irrigated.
But the Dispenser can be used with a Part Washer too. In this case, it will mix and dispense the alkaline cleaner and water emulsion, in lieu of the coolant & water emulsion in the case of a CNC Machine.

I am talking about Automatic Part Washers. Re - https://jenfab.com/
 
  • #21
Averagesupernova
Science Advisor
Gold Member
3,603
607
...concentrations & volumes as per the requirement which will vary on a day-to-day basis (not the concentration, but the volume).
So which is it?
 
  • #22
So which is it?
For a single machine, the volumes of the coolant+water emulsion required for top-up will differ, but the concentration shall mostly remain the same.

This stands true only for one machine as the coolant concentration variation shall follow a trend of whether it increases or decreases as per the parts that are being machined in that machine.

But other machines could have a completely different usage of coolant w.r.t concentration & losses per day and change in concentration (whether it increases or decreases)

if I use a Dosatron, I would have to install one in each machine, whereas with a Dispenser, I could control multiple machines from one centralized unit by incorporating some piping and valves.
 
  • #23
Averagesupernova
Science Advisor
Gold Member
3,603
607
How important is it that mixture A for machine A does not get into machine B by shared plumbing? So far this whole thing seems somewhat subjective to a bunch of unknowns to us.
 
  • Like
Likes Tom.G
  • #24
Guys,
I hope you guys are keeping well.

I have been working on developing a Compact Dispensing Station for my CNC Machine & Parts Washer where coolant is mixed with water in different concentrations & volumes as per the requirement which will vary on a day-to-day basis (not the concentration, but the volume).

My design consists of two lines with valves & Flow meters, one for coolant (pumped from a barrel/drum) and the other for water (coming directly from the plant's water supply pipes). Concentration of coolant required is in the range of 1-3% for top-up of sump & liquid volumes(mixed) required will be different everyday, as per the losses in the CNC sump. But normal range I am expecting would be around 50 Litres +- 10/20%

My system consists of a Direct Acting Solenoid Valve followed by a Flow-meter which detects liquid flow-rate & sends signal to the PLC to shut-off the valve when desired flow is met. This is in both the lines (coolant & water).

Now since a 2% coolant for 50Litres would mean approx. 1 Litre of coolant in 49 Litres of water, how do I control the flow of both the liquids so that I achieve homogeneous mixture through-out. I cannot run both the liquids together cause then how would I meter the flow of coolant to match the flow of water?

I do not have space for another tank where I can drop both these liquids and achieve mixing.

I am using this as reference - (Skip to 4:10 in the video).

Any help will be appreciated.
Thanks,
KV
I would recommend using a Pitot Venturi with a control valve to introduce the coolant into the water.The suction of the venturi will vary with the flow volume,but the relationship will not be linear.
The K factor will depend on your system variables,but once determined,it can be programmed into your PC to control the mixture.
This will allow a simplified control methodology.
I would use the suction pressure as input to the PC to control a variable valve,not on-off.
Here is a link that may help you calculate the flow/vs/pressure of the pitot tube.
 

Related Threads on Controlling two valves with different flow-rates to achieve mixing

  • Last Post
Replies
10
Views
4K
  • Last Post
Replies
3
Views
1K
Replies
5
Views
7K
  • Last Post
Replies
3
Views
2K
  • Last Post
Replies
1
Views
863
Replies
1
Views
3K
Replies
8
Views
4K
  • Last Post
Replies
4
Views
6K
Replies
1
Views
3K
  • Last Post
Replies
7
Views
3K
Top