Heat Exchanger:- Fluid Flow

To get turbulence flow, there are a number of requirements such as viscosity of the fluid.

What you do need is the calculation of a dimensionless number or dimensionless quantity called Reynolds Number.

Without giving you a straight out answer, This link will give you a start.

Engineering Toolbox

and

http://www.efunda.com/formulae/fluids/overview.cfm

I will give you more direction if you require it......when I have time.

In the mean time read up on this.

CRANE is great. Good tip

one more thing.

to answer your first question of:

but do we need turbulent flow in the tube section.

That depends, I had a micro brewery call me about 15-20 years ago, where they needed to cool or heat (I can't remember) their beer after carbonation, so they were not looking for turbulent flow.

It was a low flow rate, but I had 120' of 4" double tube. The dwell time in the tube was long, still I doubt if it was long enough, but they were willing to try it.

Simple. Get the Reynolds Number up above 2300 or so.

If the flow is laminar, the liquid layer in contact with transfer surface gets warmer and the layers away get passed off without much use. So an arrangement to facilitate best mixing would ensure better heat exchange. In shell side the baffle arrangement not only ensures good mixing but also extends the retention time to optimize the exchange. In tube side, equaling to this is multi-passing. Every time tube flows take a turn, there is good opportunity for reasonable mixing up and disturbs the laminar flow.

Besides, any steps towards creating turbulence, by adding hardware, could add up faster fouling. Since in most cases water is preferred at tube side (for easy cleaning by blasting or brushing), it could be better idea to leave tube bore bear and plain.

1. Use sufficient velocity.
2. Insert twisted ribbons or other devices into the tubes.
3. Use enhanced-surface (i.e., rifled) tubes.
4. Increase the number of fluid passes (if ΔP allows).

Number 2. is called Turbulators

As long as the OP doesn't search for turbo encabulators....

More than likely, we would have to do that for him, and then the fun will begin..........

Dear Mr.Tornado,

Referring to your Reply, I have some observations, on Point No.3 and Point4, as below.

On point No.3, Using Enhanced Surface, will give additional heat transfer and it may not give additional Turbulence.

Increasing Number of Fluid Passes - it is corolory of Point 3 will give additional heat transfer and it may not give additional Turbulence.

I want to know your comments for my views.

Thanks,

DHAYANANDHAN.S

As for 3, enhanced surface will not increase turbulence much, but it still enhances heat transfer, which is the ultimate objective.

For 4, increasing the number of passes increases the velocity, and hence the turbulence.

FLUID FLOW inside tube IS continuous mixing .Laminar Mixing lengthwise ensures THAT.

laminar flow creates boundary layers along the wall of the tubes, these do not mix and actually insulates the core of the fluid flow.

Not all White or Black

What is?

That at low Re there will not be ANY Conductive heat transfer.And that there will be no mixing inside the tube---

I understand that, what I mean is what is there that is just black and white. More things than not is compromise.

Is there any possibility of a micro helicopter mission inside the tube?

"do we need turbulent flow in the tube section"

No.

Turbulence can help break down the barrier layer which is formed due to friction to help improve heat transfer. But it is not necessary. The barrier layer thickness increases as velocity increases. The barrier layer acts as an insulator.

Are the baffles that you mention creating turbulent flow? Or are the really directing the fluid in multiple passes across the tubes. With out the baffles the fluid would travel some what in a straight line from input to out.

but the barrier layer need to be broken down. as you say it acts as an insulator... there for we need turbulence for proper mixing.

By having laminar flow, heat transfer can still take place...... but very inefficiently.

I wouldn't say you need turbulence for proper heat exchange. Heat will transfer even if you have prefectly laminar flow. In all of the tube and shell heat exchangers I have seen the tubes are quite small with significant surface area exposed to the fluid (water steam etc.) within the shell. The tube and shell are designed of sufficient size that they allow sufficient heat transfer.

As PFR said in his comment below practicality is as important as efficiency. If size is important you might spend more to have a system that induces turbulent flow to allow more heat transfer in a smaller area.

There are other types of heat exchangers that use different methods, I like the plate type heat exchanger. The surface of the plates are rippled to promote mixing and increase surface area. The flaw with these is that they need to be carefully maintained so that they don't leak.

Drew K

On the plate HX, it depends

not your product running through it has a high viscosity or has a high fouling factor.

True, you also cannot clean it with those small foam balls. They do require cleaning, and due to hard water we need to acid wash ours often.

Drew K

We run a product with a high brix, that also have a lot of contaminants in........

Maintenance requirement on this is outrageous

No, not if velocity of the fluid is kept low. Then the barrier layer will be kept thin. All this needs to be taken into consideration on the choice of or design of the heat exchanger.

In putting any device in the tubes to create turbulence. You are also creating more resistance to the flow. Slowing the fluid down. Not only that you create a large barrier layer behind the device. Like a rock in a stream and the dead calm water behind it. So you are giving up surface area in the wake of the device.

OP: - do we need turbulent flow in the tube section:?

(and do we get the micr-bubbles of air and effervesced gasses out from certain fluids, if applicable, avoiding another insulative property in slower flows)

INFO: Ernie Cherry at AquaSystemsInc.com OEM shell around coil, for fluids to DX refrigerant for too, as would

http://www.alfalaval.com/Pages/default.aspx and many others like those

IN Poly-E piping (avatar pond-coil) the deformaties of the tubing are sufficiently resulting in thermal energy transfer(txt) at 1/3- to 1/2 the 'reynolds #'s'.

3/4 i.d. "properly" transfers well-enough at just 1.5 gpm; but micro-bubbles removal may require 3.5 gpm, - and a T'd-off-line reservoir or air separators can miss receiving those in-line bubbles of collected gasses in the tubing if not flushed once in a while.

"PROPER" txt is simply what you need and may be that you need reynold's-# flow rates for smoother conforming tubing.

What's available : Heat exchangers with a "starfish" convoluting throughout the length of various depths may be ordered, as well as a very fine 'rifling' spiral of a ridge (seen.010-inch) inside the inner wall of tubing. - trade-offs with pressure drop increases, but as said above: not all is black and white

"laminar flows are not created equal". Some years ago I was researching the efficiencies of heat transfer in a spirally wound PS core air-to-air heat exchangers(Heat Recovery Unit), for use in the 200 to 600 cfm range. I found that, as we decreased the velocity of the air flow within the core, heat transfer was predictable at the low end of the turbulent scale dropping gently as velocity was reduced, and then we noticed an increase that almost rivalled that of the low end turbulent, after that peak it dropped off sharply. These tests were done initially on spirally wound sheets with flows in opposite directions. Later we tested with similar results using thin wall CAB tubing in diameters of 6mm up to 20mm and lengths of 24" to 72". However when we mirrored two chambers and so reversed the flow through the second chamber we found the most notable result was not a change in transfer efficiency, but that it was almost impossible to predict where freeze up would occur, which proves that while there is laminar flow within the tubes there was a most adequate mixing when reversing direction. When we found the ideal velocity to achieve maximum heat transfer at the minimum sound level we were able to predict similar results for units of other cfm ratings.

A few remarks:

in a heat exchanger as described the fluid with the higher convection coef is outside and the one with a lower inside but at a HIGHER velocity. For instance oil-water will be such that water is outside and oil in the tube. Of course if you look in any+even basic heat transfer book or manual you will notice that the convection coefficient goes up proportional to velocity (but not at power 1). As it was mentioned heat exchangers are designed for lowest cost of ownership which has to consider a lot of parameters. If for instance velocity is increased, for a given thermal load, dimensions will decrease and according investment but pressure losses will increase thus energy cost will go up. Increasing turbulence is a way to decrease dimensions (and cost) but could lead to an energy increase for circulation. So that it is difficult to say a priory that turbulence is a must. Further more the mentioned "turbulence generators" could be a problem with fluids which are not 100% clean and a risk of surface pollution does exist. Cleaning with such devices will be a problem. However taking all those factors into consideration it is possible to use tubes with internal extruded profiles which increase surface and turbulence but allow a not too complex cleaning. The trend however is toward higher turbulence for smaller equipment since the cost is also increased by all what is around the heat exchanger (building, aso).