The heat transfer coefficient is a very important parameter during the thermal application, and depends on many parameter such as the materials of fluid temperature. But the other important group of parameters effect on the heat transfer coefficient are flowing properties, geometry and thermal conditions.
Heat Transfer coefficient for forced convection is different than natural convection.
Forced Convection:
1. Dimensionless heat transfer coefficient is represented by Nusselt number. Nusselt number is in general function of Reynolds and Prandtl numbers. For high temperature gradient, it may become function of thermophysical property of the fluid.
Nusselt number (Nu) is heat transfer coefficient (h) multiply by reference length (L) and divided by thermal conductivity of the working fluid (k), i.e., Nu=hL/k.
Reynolds number (Re) is average velocity of flow (V) multiply by length scale (L) divided by kinematic viscosity (nu), i.e. Re=VL/nu
Prandtl number (Pr) is the ratio between kinematic viscosity (nu) divided by thermal diffusivity (a), i.e., Pr=nu/a (which is thermophyical property of the fluid).
then heat transfer coefficient is function of (k/L, Re and Pr). The function itself depends on the flow conditions (laminar or turbulent) and geometry of flow direct and geometry of the solid surface.
Natural convection:
Just replace Re by Rayleigh number ( which gravity* delta T, L^3/(nu a), all above argument is valid.
In simple words, h is complex function of fluid property (not material property), flow conditions, geometry of the flow...
Heat transfer coefficient is not a thermophysical property. Heat transfer coefficient varies with materials and flow conditions while a thermophysical property is associated with materials alone.
Thanks Prof. Junling Hu for your interested answer. So is it any classification for properties depends on materials and other properties.
Junling is right. Heat transfer coefficient is not a thermophysical property of material, while it depends on the surface geometry and the conditions of boundary layer . But the conductivity is thought to be a thermophysical property of the material. Actually the boundaries of this classification are limited to the scales of the phenomena and to the governing rules we use. Usually, materials characteristics are defined in constitutive equations as an empirical rule (e.g. Furrier's law of conduction), and these rules are not of fundamentals law. Therefore their classifications are not fundamental and are very limited to the specified scales of time, space and physics of the equations. Thus you can roughly say "the Conductivity is a thermophysical property and the Heat Transfer Coefficient is not" but you know that the conductivity is a an anisotropic tensor of forth order (with 81 component, like Elastic modulus) which, in the atomic scales, could be a function of everything happening through the material, even the heat flow!
No. Heat transfer coefficient is not a thermo-physical property. It depends on many factors like geometry, temperature of fluid, other thermo-physical properties like viscosity, thermal conductivity etc.
Junling is right. I think you are talking of convective heat transfer coefficient because from a surface to a fluid there is also a radiative one.
The former is dependent on the flow geometry and a number of parameters which inchde the thermophysical properties (thermal conductiovity, density and specific heat of the fluid) and the kiinetic ones (fluid velocity, level of turbulence and so on).
For a given flow geometry and in order to reduce the number of parameters, one describes the convective heat transfer coefficient in a dimensionless form, as either as a Nusselt or Stanton or Grashof number. Being dimensionless, these numbers may depend only on other dimensionless numbers such as the Prandtl, Reynolds and Mach numbers.
In quasi-incompressible flows, the Mach number influence is negligible.
See any classical book on convective heat transfer (e.g., Kays).
Heat Transfer coefficient for forced convection is different than natural convection.
Forced Convection:
1. Dimensionless heat transfer coefficient is represented by Nusselt number. Nusselt number is in general function of Reynolds and Prandtl numbers. For high temperature gradient, it may become function of thermophysical property of the fluid.
Nusselt number (Nu) is heat transfer coefficient (h) multiply by reference length (L) and divided by thermal conductivity of the working fluid (k), i.e., Nu=hL/k.
Reynolds number (Re) is average velocity of flow (V) multiply by length scale (L) divided by kinematic viscosity (nu), i.e. Re=VL/nu
Prandtl number (Pr) is the ratio between kinematic viscosity (nu) divided by thermal diffusivity (a), i.e., Pr=nu/a (which is thermophyical property of the fluid).
then heat transfer coefficient is function of (k/L, Re and Pr). The function itself depends on the flow conditions (laminar or turbulent) and geometry of flow direct and geometry of the solid surface.
Natural convection:
Just replace Re by Rayleigh number ( which gravity* delta T, L^3/(nu a), all above argument is valid.
In simple words, h is complex function of fluid property (not material property), flow conditions, geometry of the flow...
No. Convection heat transfer coefficient is a parameter that groups dependent and independent variables. Some of these variables are thermo-phyical properties of the fluid as k, density, viscosity,.......
The convection heat transfer coefficient, h (as opposed to the radiative heat transfer coefficient, h_r) is not a thermophysical property. It depends on the flow conditions, the geometry and whether the flow is fully developed or in the entrance region. An analogous example in fluid mechanics is the Darcy friction factor, f. The f depends on the Reynolds number, Re, and if the flow is not laminar, on the relative roughness (e/d). Depending upon the flow and surface conditions, it is possible to predict h from f (using the j_H Chilton-Colburn dimensionless factor). Hope this helps.
May i ask the same question in a different way?
If same geometry is made with two different materials but same surface finish is maintained and both are put to the same initial temperature (let 700 deg C). If air of 27 deg C is blown with a rate of 1kg/sec over these two. Whether heat transfer coefficient will remain same or not?
@Naveen Bhatt, As the heat transfer coefficient in forced convection depends on Nusselt number (depends on geometry, irrespective of materials & thermal conductivity of gas), Reynolds Number (dimension, viscosity, Velocity & density of fluid) and Prandtl number (on viscosity, specific heat, thermal conductivity of fluid).
Now, in your case, geometry is same, same temperature of same fluid, velocity or mass flow rate is also same, Thus, h will be same for same surface finish body in your case..
That means for a specific heat exchanger, with different streams we have different heat transfer coefficient?
q = UA.deltaT, in the retrofit case, can I keep UA is same in different configurations for a fair comparison?
where can i get air and carbon-ceramic matrix interface convection coefficient value?
Hi all
I confronted a basic concept!
In a case in which "h (Convective Coeff.)" only depends on hydrodynamics (i.e. Flow Field, Re) and Physical Properties (i.e. Pr), is it necessary to solve Energy eq. to get the "h."
h = a Re^b Pr^c
can one see any parameter depending on heat transfer parameters itself at the right hand side of above eq.
If so, we expect to gain the convective coeff. without the trouble of solving energy eq
heat transfer is boundary phenomenon, you need to know temperature gradient at the boundary, h is proportional with negative slop of temp. gradient at the boundary. It is correlated as you mentioned. Correlation is not definition and valid for specific geometry, flow conditions and thermophysical properties.....
Mr. Saeed Azimibavil ,
h is a dimensional quantity and it cannot depend only on dimensionless numbers. Where are its dimensions ?
Best regards
Of course h is dimensional parameter (W/m2.K). However, Nussel number, Nu, is dimensionless parameter, Nu=hL/k, then h=Nu.K/L. K=thermal conductivity and L is characteristic length scale.
- Badges
- Science topic
- Similar topics
- Engineering
- Thermal Engineering
- Thermofluid
- Heat Transfer