Forced convection - parallel flow over a flat plate - local heat transfer coefficient

[rpbyc]

Hi,
I have the following problem: how do I determine the local heat
transfer coefficient
for a parallel air flow over a flat plate, but near the leading edge?

From the books I've looked into it turns out that it's value
theoretically rises
to infinity near the leading edge, which is impossible (I think) in
practice.

Thanks in advance for the help...

--
Regards,
Robert

[me]

On Wed, 21 Oct 2009 00:24:13 -0700 (PDT), rpbyc <rpb...@gmail.com>
wrote:

>Hi,
>I have the following problem: how do I determine the local heat
>transfer coefficient
>for a parallel air flow over a flat plate, but near the leading edge?
>
>From the books I've looked into it turns out that it's value
>theoretically rises
>to infinity near the leading edge, which is impossible (I think) in
>practice.

Yes the local heat transfer coefficient goes as 1/sqrt(x). But these
are classic solutions for both laminar and trubulent flat plate BLs.
Maybe if you explain your need a bit more, ie are you interested in
the overal heat load over some portion near the LE, or how close to
the LE are you talking about?

[Robert B.]

On 21 Paź, 21:06, me <m...@mine.net> wrote:

> Yes the local heat transfer coefficient goes as 1/sqrt(x). But these
> are classic solutions for both laminar and trubulent flat plate BLs.
> Maybe if you explain your need a bit more, ie are you interested in
> the overal heat load over some portion near the LE, or how close to
> the LE are you talking about?

I have a 7 cm long aluminium radiator, attached to a power transistor.
The power dissipated
by the transistor is, at present, 30 Watts (constant power).

The question is: can I assume that the convective heat transfer
coefficient is constant over the
whole surface of the radiator?

The radiator is cooled by an air vent, which is at a distance of about
6 cm from the
leading edge of it. The velocity of the air stream, measured just
before entering the leading edge is
about 2,4 m/s.

Regards,
Robert

[Robert B.]

On 23 Paź, 13:58, m...@mine.net wrote:
> On Fri, 23 Oct 2009 02:20:01 -0700 (PDT), in
> sci.physics.computational.fluid-dynamics "Robert B."

>
> <rpb...@gmail.com> wrote:
> >I have a 7 cm long aluminium radiator, attached to a power transistor.
> >The power dissipated
> >by the transistor is, at present, 30 Watts (constant power).
>

> If this is a commercial heat sink, I would think you would be able to
> find data for its specific shape.

I haven't look for the data yet, but I'll try to.

> >The question is: can I assume that the convective heat transfer
> >coefficient is constant over the
> >whole surface of the radiator?
>

> No, but you could integrate the expression for the local h_c over the
> length and evaluate that to determine an average hbar_c.

>
> >The radiator is cooled by an air vent, which is at a distance of about
> >6 cm from the
> >leading edge of it. The velocity of the air stream, measured just
> >before entering the leading edge is
> >about 2,4 m/s.
>

> Assuming air at 20 deg C:
>
>         nu = 15.7E-06 m^2/s
>         k  = 0.0251 W/m-K
>         Pr = 0.71
>
> For a flat plate with L = 0.07m and U = 2.4 m/s
>
>         Re_L =  U*L/nu  = (2.4 m/s)*(0.07m) / (15.7E-06 m^2/s)
>
>         Re_L = 1.07E4  >> laminar BL
>
> so,
>
>         Nu_x = h_cx * x / k = 0.33 * Re_x^0.5 * Pr^0.333
>
> integrate this from  x = 0 to x= L.

Thanks very much for your help. I've had already done the above
calculations. It looks I should
be able to apply the mean value. The measurements show that the
temperature difference between
the leading and trailing edges is not higher than 1K, whereas the
overall temperature of the radiator's surface
is about 45 deg. C.
So that in my application I think that the mean value of h_x should be
fine.

Regards,
Robert