Flowing water temperature mixing

[Valalvax]

Useless explanation text:

I'm building a shower system that uses motors to manually turn valves to adjust the temperature based on what the user sets

I'm trying to develop a way for the system to preemptively predict what the temperature will be so it doesn't just say "Open hot valve a little more, nope, that's too hot, lower hot valve and raise cold valve a little... no that didn't work"

More important information starts here:

With it being a shower obviously the temperature is mixed, and is there is only a short time until it exits via the nozzle (so we need to account for time, or don't, this isn't really meant to be ultra precise, call it 1.5 seconds and we're good)

Poking around I found the formulas for calorimeters and such, but that assumed you poor source A and source B in, and waited until it was thoroughly mixed before measuring, whereas in my situation the water is flowing, not in a container, and isn't necessarily given the time to completely mix

Quick random example:

Hot Water: 115 degrees Fahrenheit at 1.25 G/Min
Cold Water: 40 degrees Fahrenheit at 1.25 G/Min
Time in System: 1.5 Seconds
Heat Specific of Water: 4182

Anyway, sorry for being so wordy, better to give too much information than not enough...

 

[Penn.6-5000]

The time in the system is irrelevant; all that matters to us is that the mixing is thorough. And that's not all! If you're just mixing the water but not adding any extra heat, you can just take a weighted average!

1 gallon of 115-degree water + 1 gallon of 40-degree water = 2 gallons of 77.5-degree water.

Suppose you change the mix to 85% hot, 15% cold. The final temperature is:
0.85*115 + 0.15*40 = 102.75 degrees.

The same rule works even if the water is just flowing through the system. If 85% of the water coming in is at 115 degrees and 15% is at 40 degrees, it doesn't matter how much water passes through the system each second or how long it spends in the system. The output will be 102.75 degrees, and you'll have as much water coming out as you put in.

 

[Valalvax]

But won't it require some time to mix thoroughly? (Though I suppose movement will speed the process greatly)

And what about all the fancy heat specific stuff for calorimeter's? Surely it matters (or not as much because they're equal anyway)

 

[Penn.6-5000]

It does require some time to mix thoroughly and yes, the fact that it's moving will help it mix pretty quickly. What I meant was that the fact that the exact amount of time is of no consequence. There's no equation into which you can plug "1.5 seconds" to get a useful answer. But it's reasonable to say that if you don't have enough time for the water to mix, then some of the little streams coming out of the showerhead will be warmer than others.

Specific heat (not 'heat specific') is a measure of how much energy it takes to cause a certain amount of a material to increase in temperature by one degree. It would matter if you added an electric heater to your shower system. How many watts of heat would it take to cause 1.5 gallons of water per second to warm up by 10 degrees? That's a specific heat question. You also need specific heat when heat energy moves from one substance to another. If you drop a pound of 40-degree copper into a pound of 60-degree water, the final temperature isn't 50 degrees; it's closer to 58. The reason is that it takes more than ten times as much energy to heat a pound of water by 1 degree than it does to heat a pound of copper. So when enough heat leaves the water to make it one degree cooler, the copper gets ten degrees hotter. How do I know that it's ten times as much and not, say, fifty? Because I compared the specific heats of water and copper.

So all that fancy specific heat stuff matters for more complicated problems, but when heat moves from one bit of water into another bit of water, a simple averaging argument is sufficient.

 

[Valalvax]

Thanks, and meant specific heat... serves me right for trying to guess at something I'd read like four hours ago without rechecking it :/

Thanks for the help, now I have no excuses to delay this project anymore *grumble*

 

[S_Happens]

I know it's not quite the question you asked, but it may make your project easier.

The easiest, and least complicated to program/control, would be a single valve that mixes the hot and cold water, similar to most stand up shower controls. This gives you one temperature reading (downstream of this valve) and one control point for a very simple control loop.

Trying to control two valves independantly, especially based on one temperature reading will be much more difficult (and really a waste of time). Would you use two controllers each reading the same temperature, or try to control two valves with one controller? No matter, it would be difficult to prevent the two from "fighting" each other, as there would be double moves made. Typically the way to use two vavles and one controller is to end up manually (also if you're going to automate control of the valves, then mechanically operating them in the control loop isn't "manual," but automatic) setting one valve in a certain position, and automatically controlling the other with the control loop. One variable, one control point, easy. Another way would be to open the valves in series, one after the other (split range), although that wouldn't ve a very good way to do it in this case.

Back to my suggestion. Use a single handle mixing valve, very commonly used for showers and baths. This means one variable and one control point. It will mix the water for you, and make the control loop very easy. A temperature increase will call for one direction, while a decrease would call for the other.

How do you plan on controlling this?

 

[Valalvax]

Two manual valves that are turned by stepper motors behind the wall, I'll probably use a self-made coupler to couple them together, and an Arduino will control the whole mess, the temp gauges will be at the shower head and immediately before the hot and cold water valves.

Why two valves rather than one valve? No particular reason other than "I want to."

The actual user interface (which I doubt you care about) is still undetermined, I'd love to use a LCD screen, but they're expensive, another idea is using 27-segment LEDs for text and stuff..

The starting menu will have everyone's name, and a guest account, if you log in then start the shower it'll immediately go to your average temperature and after it reaches the temperature it'll inform you it's ready (the guest account is preset for 105 degrees, based on an average preferred shower temperature I found online), I'm toying with the idea of it warning you that it's becoming increasingly impossible to hold your temperature, maybe even use a flow meter to determine that the hot water tank is nearly empty (with the knowledge of how much it holds of course)


Anyway, I'm rambling at this point, I'm very much a "Get this working, then think about that" kind of person so a lot of this isn't 100% sure in my head yet (and that's really a bad thing because if "that" requires changes to "this" then I need to start over practically)

 

[S_Happens]

OK, so you're modifying some manual valves with your own actuators. Do you know what type of valve they are?

Why two valves rather than one valve? No particular reason other than "I want to."

You do realize that this tends to make actual temperature control difficult. Like I said earlier, depending on how you ultimately decide to control each valve (independantly or together) the tendency will be to fight each other and/or make double adjustments. This is compounded by the fact that you're controlling temperature, in which your measurment will lag. You may be able to have some success if you use very slow control, but I think you'll see cycling.

My familiarity isn't at the hobbyist scale, but at very large budget manufacturing scale, so I do not know what kind of software is available to you. My suggestions are coming from dealing with software and equipment from very large companies dedicated to process control.

From what I see you asking/posting, it looks like you want to use 3 temperature sensors and control two valves. This is quite a bit more complicated than neccessary, which is why I have made my suggestions. Using 1 temperature sensor downstream after the hot and cold combine and controlling a single valve would be much simpler AND more effective (not necessary to make any calculations based on the temperatures of incoming hot and cold water). I guess my question is, are you just wanting to accomplish the goal and looking for practical advice or are you set in the way that you're going to go about this (2 valves and 3 temperature sensors)?

Edit- Very interesting project BTW.

 

[S3196655]

Hey Valalvax
I am currently undertaking a similar project also. Have you had any luck with a mixing valve. I have looked at all sorts of 3 way rotary mixing valves controlled by an actuator. These all seem very expensive. Was wondering if you found a cheaper solution.

 

[Valalvax]

The project never wound up happening, I have a bit of code written, but I kept making it more advanced and more advanced, and never touched the hardware aspect of it, and found out we were moving this August, so I figured it wouldn't be worth it to extensively mod a house we weren't staying in, so I left it for now.. the way I was going to control the valves was through a stepper or servo motor, seemed a bit cheaper than a automatic valve

 

[S3196655]

Thats a shame. Hopefully I get my project finished since its for my end of year final submission for Uni. Did you have any luck sourcing a valve though could be controlled by a stepper motor? Have seen one designed by Aqualisa (http://track-days.co.uk/gassing/top...t=Aqualisa+Axis+Digital+-+naked+photos!&mid=0) and Moen (https://www.amazon.com/dp/B004S654QC/?tag=pfamazon01-20) but they all seem expensive.

 

[Valalvax]

You misunderstood, I meant buy a valve, buy a servomotor... and... my schooling is failing me right now.. (can't think of the words) but... attach the rotor of the motor to the stem/handle of the valve, write code to control the opening/closing of the valve...

 

[S3196655]

ahhh. seems like a legit way... and a whole lot cheaper. really baffles me why these valves with stepper motors combined are so expensive.

 

[Valalvax]

ahhh. seems like a legit way... and a whole lot cheaper. really baffles me why these valves with stepper motors combined are so expensive.

Couple reasons: They're bound to be 10 times more precise than anything you and I can cobble together

They have to apply for a lot of licenses and stuff to make sure it won't blow up/catch on fire etc etc

It's really expensive testing and coming out with this kind of stuff on a commercial level