Heat transfer coefficient calculation ( Cooling system efficiency issues) 5

[Annelissenc.]

Hello and thank you for having me in this forum.

I am trying to calculate the heat transfer coefficient of an open square etching tank containing 5000 L diluted 1% HF and 4% H3PO4 water solution. This tank has been immersed in a outer tank covered with tap water. The outer tank contains water, burner tubes and a cooling system to keep the temperatures between 41 and 46C. the inner tank is used to etch Aluminium and the by-products are aluminium fluoride and hydrogen. In a period of 5 weeks, the build up of aluminium fluoride in the inner tank can go up to 2500kg and creates a thick layer in the walls that works as an insulant reducing the efficiency of the temperature exchanges. Temperature goes above the etching limits (41-46C) and it is difficult to bring back down heating up to 50+C and taking 3+ hours to reach back operational temperatures. This is an issue that gets worse towards the end of the bath life (5 week old etching acid) and affects production speed.
I started an experiment where I am recording the temperatures of the inner and outer tanks before and after etching to see if over time the differential increases and I would like to be able to measure the heat transfer coefficient over the 5 weeks to experimentally see when and how this build-up is affecting the cooling and burning system efficiency. Can someone help? Suggestions and other approaches are welcomed
Materia Stainless still 316L grade
L =(Thickness ) = 3mm ( to be measured to confirm)
Iner tank leght = 1.900m
Iner tank with = 0.862m
Iner tank heigh = 1.903m
Inner tank A = 12.5m2
Cooling system uses Glycol as coolant
T1 and T2 experimentally

 

[EdStainless]

First, nice post, but please don't use all bold as it actually makes it difficult to read.
It is likely easier to get good numbers for the cooling case.
You know the amounts of materials (volumes and masses) and their thermal capacity so you can work with delta Ts and times to get coef.
You might just want to record time to cool from 50C to 40C (or some such range) at different points during the 5-week cycle.
Use the fresh clean tank as a baseline and evaluate for the decrease in heat transfer.
The problem with measuring on heating is that you would need to know the exact heat input.
What your case tells me is that operationally 5 weeks is too long.
More likely you should be tracking the weight (or surface area) of parts processed rather than overall time.
In many cases people partially clean and refresh tanks more frequently.
If you we were to change half of you tank every two weeks and clean part of it out then you could operate for a longer period with less variation in performance (and results).

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[georgeverghese]

Try to find a chelate that will keep AlF3 in solution which will eliminate scaling the inner tank walls - Google tells me triazo cyclononane may be one choice. Also speak to chemists in water treatment companies - for example, they have chelates that keep calcium, magnesium and barium in aqueous solutions in the dissolved state - commonly called antiscale agent.

 

[Annelissenc.]

post 1
Sorry for using bold, it might have just passed, it was nonintentional. Thank you for your interesting answer. Ideally reducing the cycle time would be a good idea, but this is a quite unique process and the aluminum in solution actually helps to stabilize the process we need a minimum of 15000 ppm of aluminum in solution to start the tank ,for that, in the 2 tanks we have we need to keep a dilution balance during those 5 weeks production cycle. On the other hand, we are talking about 5000 L tank1 and 12000 l capacity, the precipitation (Aluminum fluoride) formed is rock solid so we need 2 days to clear them out with a proper pneumatic drill.
The solid build up depends on the material etched and goes from 0 to 2.600tons, from 3.000 we lose control of the process. The weight approach is interesting, and I Know exactly how much weight we build up through time by looking at the etched material, it would be interesting to corelate that with the increasing of cooling times. The build up of aluminum is not homogeneous in the tank which makes it challenging to gestimate the formed insulation.


for post 2
Because of the dangerous nature of the system we normally decide against adding new chemicals, especially if these chemicals are reactive, the whole process is designed in terms of etch rate and any change in chemical reactions can change those rates... If however, these chelates solve the build up problem, we are happy to explore this possibility.

 

[georgeverghese]

Low heat transfer coeff on the surface of the inner tank may be the reason for this buildup. With low htc, you need high surface temp, and with that, you get AlF3 dropping out of solution. Check the solubility vs temp curve for AlF3 - does solubility decrease with increasing temp?
And why do you let AlF3 concentration continue to buildup to beyond its solubility limit?
Is Al2 (SO4)3 perhaps more soluble than AlF3?

 

[pierreick]

Hi,
I believe this is the consequence of the chemistry, your solution is over saturated, and materials are precipitating over the time. Aluminum phosphate is not soluble.
NB: You may try lab tests to define the optimum condition versus time before replacing the solution. Ask support of a chemist within your organization.
My 2 cents
Pierre

 

[LittleInch]

Avoiding the chemistry issue and looking at this from a heat transfer issue, can you sketch or describe your system some more.

So how thick is the water layer between inside and outer tank
Is the outer tank insulated over all six sides?
Is the water circulated to promote homogenous temperature?

Could you add fins to the outer side of the inner tank?

If your issue is that the inner tank gets too hot why are you restricted to only 41C in the outer tank?

But if you assume no or very little heat transfer between outer tank and the ambient air, then your heat flow is measured by either the energy in through the heating or energy out via the cooling system ( Delta T x flow rate x Cv) This only applies for a steady state situation when temperatures are not moving.

Why are you using glycol for your cooling system? What percent? what temperature is it operating at.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Annelissenc.]

georgeverghese / pierreick

I partially agree with you, and yet is not that simple due to the etching rates and specs. The operational temperature in the inner tank has to be between 41C and 44C and the etching times need to be controlled. if we go above this temperature to increase solubility we risk over etching (or flash fires when it goes above 50C depending on acid concentrations), this without counting that we have an exothermic reaction so we generate a lot of heat .

At this temperature range our saturation point goes anything between 60000ppm and 70000ppm aluminum content. These concentrations can be reach after 3-week production, so unfortunately, we have to live with these precipitations, or we would be constantly cleaning the tank involving other Including cleaning cost which for these dimensions is Hight.

Note that the inner tank contains HF 1%in solution and H3PO4 4% in solution (phosphoric to take the oxide layer of aluminum and HF for actual etching) and we etch aluminum blocks, therefore Aluminum fluoride is the result of the etching process, a by-product.

We are considering moving for an alkali etching method but is not going to happen any sooner so this is what we have to work with at the moment.

 

[Annelissenc.]

LittleInch


1 - So how thick is the water layer between inside and outer tank –
About 0.5 m
Inner tank dimensions: l=1.9m, w-0.862m, h-1.903m

2 - Is the outer tank insulated over all six sides?
it has 5 sides, the tank is open, it is stainless still 362L 4mm and is not insulated. Neither is the outer tank, he build-up over the time inside the etching tank (inner) creates the insulation.

3 - Is the water circulated to promote homogenous temperature?
The cooling system is always in circulation, but the design of the tanks does not permit agitation.

4 - Could you add fins to the outer side of the inner tank?
How does this work? We have the burners in the left-hand side between the tank for when the temperature drops to 41 or below.

5 - If your issue is that the inner tank gets too hot why are you restricted to only 41C in the outer tank?
If the inner tank goes below 41 C the burners in the outer tank are automatically activated (I might have not explained this week)

6 - Why are you using glycol for your cooling system? What percent? what temperature is it operating at.
I believe is poor design that I have been left as a legacy … the cooling tank has a volume of 2m3 and contain 25% glycol and 75% water. The temperature it operates is ambient dependant as the coolant circulates to a 3-fan system (sat outdoors) that cools down to 32C, this process in summer is very inefficient as it takes a lifetime to cool the tank.

 

[LittleInch]

Anne... Can you post a sketch / cross section as I'm confused now.

If the outer tank is not insulated, then measuring the heat lost by the water is going to get complicated as some will escape via the tank walls and some via the cooling system. ditto in any heating mode some heat will be lost to the ambient air so it makes calculating the heat transfer from the inner tank to the outer tank difficult and hence you won't be able to calculate OHTC to any degree of accuracy.

Need a sketch, but if you increase the surface area of the inner tank to the water jacket, you should get better heat transfer. Burner tubes are not usually that big.

Can't you just adjust the temperature the burners come on when you're in cooling mode or just turn them off when not needed?

If your water is not subject to freezing conditions then just reduce the glycol content to virtually nothing as you'll get much better heat transfer with water alone.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jari001]

I think you attached a different file than a sketch. You may want to remove this content ASAP.

 

[Annelissenc.]

jari001 (Chemical

Thank you...I have reported it as I cant delete it

hope this time the link is correct.... it keeps happening

 

[IRstuff]

Given the lack of apparent symmetry and uniformity of the cooling/heating/etching environments, doing a calculation would seem to be difficult, at best.

You've not shown whether the cooling/heating system has any means of uniformly heating the tank, so there could be large temperature differences between the 5 walls of the tank.

The etchant itself, without agitation, would likewise have potentially large thermal gradients, even ignoring whatever thermal impact the etching process itself has.

What is the size of the object being etched, relative to the tank size?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[georgeverghese]

Acetic acid would keep the aluminum in solution and prevent scaling.

 

[Annelissenc.]

IRstuff (Aerospace)

yes,
It is a tricky one. the recirculation system is the only way we have of agitating the tank due to space issues. The etching baskets are pretty much the size of the inner tank so a lot of heat generation, we cannot agitate the tank during etching, one because of space two it changes the rates too. I do know by fact that the heating is not all homogeneous, specially with the build up being different each time. We have had a quote for creating a whole automated clean room with controlled environment but it totally out of budget.

It seem that this problem hasn't got a straight forward solution so I am going to give up on calculations and find another way to remediate the cooling problem

georgeverghese (Chemical)

this would be a good experiment to look at, even if it delays the saturation point it would help.