Simultaneous Prediction of Solid stress, Heat transfer and Fluid flow by a Single Algorithm

By Brian Spalding

Lecture presented at XIII School-Seminar of Young Scientists and Specialists under the leadership of the Academician, Professor A.I.Leontiev

May 20-25, 2001, Saint Petersburg, Russia

next or contents












Abstract

next or back












Contents

  1. The problem
    1. Its essential nature
    2. Practical occurrence
    3. The conventional solution
    4. A better solution
  2. A multi-physics example
    1. Stresses resulting from radiation, conduction and convection
    2. Vector and contour plots
    3. How the stress calculations were performed

    next or back or contents











  3. The mathematics of the method
    1. Similarities between the equations for displacement and velocity
    2. Deduction of the associated stresses and strains
    3. The "SIMPLE" algorithm for the computation of displacements
    4. More details of the equations
  4. Details of the auxiliary models
    1. IMMERSOL, for radiation
    2. WGAP, WDIS and LTLS, for radiation and turbulence
    3. LVEL, for turbulence
  5. Conclusions
  6. References

next or back or contents











1. The problem

(a) Its essential nature

It is frequently required to simulate fluid-flow and heat-transfer processes in and around solids which are, partly as a consequence of the flow, subject to thermal and mechanical stresses.

Often, indeed, it is the stresses which are of major concern, while the fluid and heat flows are of only secondary interest.

next, back or contents











(b) Practical occurrence

Engineering examples of fluid/heat/stress interactions include:

next, back or contents











(c) The conventional solution

It has been customary for two computer codes to be used for the solution of such problems, one for the fluid flow and the other for the stresses

Iterative interaction between the two codes is then employed, often with considerable inconvenience.

next, back or contents











(d) A better solution

It is however possible for fluid flow, heat flow and solid deformation, and the interactions between them, all to be calculated at the same time.

The method of doing so exploits the similarity between the equations governing velocity (in fluids) and those governing displacement (in solids).

In the present lecture, the results of such a calculation will be shown first.

The explanation of how it was conducted will then follow.

next, back or contents












2. A multi-physics example

(a) Description: The task is to calculate the stresses in radiation-heated solids cooled by air.


    20 deg C| air

            |             80 deg C

          | V |/////// hot radiating wall ///////////|

          |   ----------------------------------------

          |                       duct              ----->  exit

          |-------------                 -------------

          |// steel ///|     cavity      |/// steel /|

          |------------------------------------------- ? temperature ?

          |////////////// aluminium /////////////////|

          |-------------------------------------------
next, back or contents











Details of the calculation are:

  1. The Reynolds number (based on the inflow velocity and horizontal duct width) is 1000.
    Therefore the LVEL model is used for simulation of the turbulence.

  2. The radiative heat transfer is represented by the conduction-type IMMERSOL model, which is:
    1. economical and
    2. fairly accurate
    for such situations.

    The absorptivity of the air is taken as 0.01 per meter;
    the scattering coefficient as 0.0;
    and the solid surface emissivity as 0.9 .

    next, back or contents











  3. Both LVEL and IMMERSOL make use of the distributions of:
    1. distance from the wall (WDIS) and
    2. distance between walls (WGAP), both of which are calculated by solving a scalar equation for the
    3. LTLS variable.

  4. The stresses within the metals result primarily from the differences in their thermal-expansion coefficients. namely:

next, back or contents











(b) Vector and contour plots

next, back or contents












(c) How the stresses were calculated

next, back or contents












3. The mathematics of the method

(a) Similarities between the equations for displacement and velocity

The similarities already referred to are here described for only one cartesian direction, x; but they prevail for all three directions.

next, back or contents











  1. The x-direction displacement, U, obeys the equation:
    where:

    next, back or contents











next, back or contents











Notes:

  1. The two equations are now set one below the other, so that they can be easily compared:

  2. The equations can thus be seen to become identical if:

    next, back or contents











  3. The expressions for C1, C2 and C3 are:


    next, back or contents











  4. A solution procedure designed for computing velocities will therefore in fact compute the displacements if:

    1. the convection terms are set to zero within the solid region: and

    2. the linear relation between:
      • D ( ie [d/dx]* U + ...) and
      • p
      is introduced by inclusion of a pressure- and temperature-dependent "mass-source" term.

next, back or contents











(b) Deduction of the associated stresses and strains

The strains (ie extensions ex, ey and ez) are obtained from differentiation of the computed displacements.

Thus:

ex = [d/dx]* U

ey = [d/dx]* V

ez = [d/dx]* W

next, back or contents











Then the corresponding:

are obtained from the strains by way of equations such as:


sx = {YM / (1 - PR**2)} * {ex + PR*ey} and

tauxy = {YM / (1 - PR**2)} * {0.5 * (1 - PR)*gamxy}

where:

next, back or contents











(c) The "SIMPLE" algorithm for the computation of displacements

PHOENICS employs (a variant of) the "SIMPLE" algorithm of Patankar & Spalding (1972) for computing velocities from pressures, under a mass-conservation constraint.

Its essential features are:

next, back or contents











All that it is necessary to do, in order to solve for displacements simultaneously, is, in solid regions, to treat the dilatation B as the mass-source error and to ensure that p obeys the above linear relation to it.

Therefore a CFD code based on SIMPLE can be made to solve the displacement equations by:

  1. eliminating the convection terms (ie setting Re low); and
  2. making D linearly dependent on p and temperatureT.

The "staggered grid" used as the default in PHOENICS proves to be extremely convenient for solid-displacement analysis; for the velocities and displacements are stored at exactly the right places in relation to p.


next, back or contents











4. Details of the auxiliary models

(a) IMMERSOL: summary

next, back or contents











Notes:

Click here for more information.

next, back or contents











(b) WGAP, WDIS and LTLS

next, back or contents











Notes:

Click here for more information.

next, back or contents











(c) LVEL: summary

next, back or contents











Notes:

Click here for more information.

next, back or contents











Click here for an SFT example involving natural convection

next, back or contents











5. Conclusions

The following conclusions appear to be justified:
  1. Simultaneous simulation of solid-stress, heat transfer and fluid flow is indeed practicable and economical.

  2. As compared with the alternative, namely the use of distinct methods for each phenomenon with iterative interaction between them, the simultaneous-solution method is very attractive.

  3. It therefore seems possible that, when its attractiveness is fully recognised, SFT (i.e. Solid-Fluid-Thermal) analysis may become as popular as CFD.

    next, back or contents











  4. However, because older specialists have too-long believed the two-distinct-method approach to be the only practicable one, the future of the simultaneous-solution approach depends on its adoption by younger ones.

  5. This is why it has been presented to the "School-Seminar for Young Scientists".

    next, back or contents











  6. Such scientists, before committing themselves to this line of research should ask: ------------------------ END of LECTURE ------------------------

    next, back or contents












6. References

  1. The differential equations governing displacements, stresses and strains in elastic solids of non-uniform temperature can be found in numerous textbooks, for example:

    It has not been common to choose the displacements as the dependent variables in numerical-solution procedures. However, this has been done by:

    Their numerical procedure differ from that used here, which was that of

  2. The first use of the present method for solving the solid-displacements and fluid-velocity equations simultaneously appears to have been made by CHAM, late in 1990.

    Reports describing the early work include:

    From that time onwards, the solid-stress option was made available as a (little-advertised) option in successive issues of PHOENICS,

  3. Open-literature and conference publications have been few, but include: