>
> Greg,
>
> I'm glad you ask the question because it points out the source of a lot
> of misunderstandings arising out of "intuitive" thinking. The human mind
> tends to look at things in a linear fashion which works fine as an
> approximation for most every day problems. However, in nature physical
> phenomenae tend to be non-linear.
=====
Agreed.
=====
Such is the case with heat transfer.
> As has been pointed out in a couple of posts by Mike and/or Candy as
> coolant passes across a hot surface its temperature increases as heat is
> removed from the source. Mathematically heat transfer for a fluid
> passing over a heat source (in its simplest expression) is proportional
> to the product of a heat transfer coefficient (a function of the material
> properties of the interface), the transfer unit area, the temperature
> difference between the heat source and the coolant, and the fluid flow
> rate. That is, Q is proportional to: Cp x A x Mdot x (Ts-Tc). This
> formula clearly shows that heat transfer increases as flow velocity
> increases.
=====
I'll have to take your word for it, since I couldn't find anything in a
quick search to either confirm or refute it.
=====
To help understand this envision your copper tube as being
> heated and having a constant temperature along its length. Now envision
> a flow of cool water into the tube. At the entry end of the tube the
> coolant has a low temperature and the tube is hot.
=====
Shouldn't this be the other way around?
=====
Heat transfer per
> unit mass of coolant is at its greatest at this point since the
> temperature difference is at its widest margin. However, as the coolant
> absorbs heat its temperature increases, thus as it flows down the tube
> the rate of heat transfer progressively decreases. The total heat
> transfer will be the product of the change in temperature of the flowing
> fluid, its heat capacity and the mass of the fluid.
=====
Agreed.
=====
If we slow down the
> flow rate, more heat per unit mass is absorbed at the beginning but the
> rate of heat transfer down the length of the tube decreases, because the
> temperature difference is smaller, and the quantity of mass flowing is
> less because the velocity is lower. Therefore we may be maximizing the
> amount of heat absorbed by the fluid but we are decreasing the amount of
> heat transferred from the source. Now conversely imagine the velocity
> of the fluid increasing. Two factors are working to increase the heat
> transfer rate; First the temperature difference throughout the length of
> the tube is increased, and secondly the quantity of mass absorbing heat
> per unit time is increased. The product of the two factors dramatically
> increases the total heat transfer rate.
>
> That being said let me point out that you are correct in noting that
> other factors affect heat transfer and you specifically pointed out pump
> cavitation. In fact there are many important factors and the mathematics
> for describing real world heat transfer problems can become quite
> complicated. Factors such as coolant physical properties, system
> pressure, geometry, surface roughness, turbulence, fouling etc. all
> effect the heat transfer. Nevertheless, the fundamental property of heat
> transfer increasing with flow rate is pretty much a hard and fast rule in
> mother nature. Pump cavitation (or cavitation anywhere in the system)
> while it can dramatically decrease heat transfer is not a violation of
> this.
=====
This is where it falls apart for me. The motor's idling at rest, There's a
~fixed amount of air moving through the radiator. With a regulator installed
(t'stat) flow is at least partially restricted, even if wide open. Without a
t'stat, coolant can flow at whatever the maximum for the system at rest
allows. This should be faster than with a t'stat. Why then does the motor
overheat without a t'stat WRT when it has a t'stat? Also, as RPM's increase,
so does coolant flow. Why does it run cooler with a t'stat WRT none?
If I understand you correctly, then the opposite should be true. What am I
missing?
=====
Rather cavitation is introducing a change in the thermophysical
> property of the coolant as it is no longer a liquid but a vapor.
=====
Agreed.
=====
>
> Understanding the nature of heat transfer is important as it can help us
> avoid potential pitfalls. For example, someone suggested using an
> underdrive pulley as a possible solution to an overheating problem.
=====
I did.
=====
In
> point of fact this will almost certainly reduce the heat transfer rate
> and worsen the cooling problem. Underdrive pulleys are frequently used
> to reduce the horsepower consumed by the coolant pump in performance
> cars, but, they can only be successfully used in cars where the cooling
> system has enough excess capacity to still cool the car with its cooling
> efficiency reduced or in situations where the vehicle will only be
> running for a short time, as in drag racing.
=====
Works well in road racing, and is a common mod with hot rodders, so a
typical stock cooling system must be really overkill. Seems hard to believe
considering how hot they get without a t'stat.
=====
>
> I hope this helps. If anyone still has questions or is interested in
> other aspects heat transfer I will be glad to respond.
=====
Not yet, see above.
(
GM
>
> Andrew
|