Thermal And Fluid Models in the Virtual Test Bed

1
Thermal And Fluid Models in the Virtual Test Bed
VTB Annual Review July 16-17, 2002
Greg Anderson Naval Surface Warfare Center -
Philadelphia, PA
2
Introduction

• Why use VTB?
• Why develop models?
• Models developed for old release.
• Advantages in new release.
• Future directions

3
Why use VTB?

• One goal is to test control logic for system
  reconfiguration algorithms
• Requires software with a true transient
  calculation engine.
• Requires software with fluid piping models
  (pipes, pumps, tanks, valves, etc) integrated
  with control logic models (Simulink)

4
Why use VTB?

• Another goal is to model of power electronics and
  cooling schemes.
• VTB handles the electronics modeling already
• Requires integration of electronics models and
  thermal models.
• Modeling the whole system requires integration of
  thermal models with fluid models.

5
Why develop models?

• For fluids, the existing models simply equated
  conceptually equivalent devices
• Pipes are the same as flow resistors
• Pumps are the same as transformers
• For thermal phenomena, no specific models exist.

6
Why develop models?

• Why not treat pipes as resistors?
• For laminar flow (low flow rates or small pipes)
  the pressure drop is linear with flow rate
  (constant resistance).
• For turbulent or transitional flow (encountered
  far more often in practice), the pressure drop is
  based on the square of the flow rate (variable
  resistance)

7
Why develop models?

• Why not treat pipes as resistors?
• Pipes can act as pressure sources when the inlet
  and outlet elevations are different.
• When one end of a pipe discharges to air, reverse
  flow isnt possible.

8
Why develop models?

• Okay, so why arent pumps just transformers?
• The voltage output of a transformer is a function
  of construction and the input voltage.
• The pressure output of a pump varies with the
  flow conditions based on the pump curve.

9
Why develop models?

• Okay, so why arent pumps just transformers?
• Transformers ideally conserve energy the input
  power equals the output power
• Pumps add power to the flow the flow leaves with
  a higher stagnation pressure than when it entered
• (stagnation pressure is a measure of energy its
  the pressure the flow would have - based on
  Bernollis equation - if it were motionless)

10
Why develop models?
11
Why develop models?

• Any other issues?
• Real tanks provide variable pressure the
  pressure available changes as the fluid level
  rises or falls because of the flow in or out.
• The tank model assumes an infinitely large
  surface area, such that the pressure remains
  constant independent of flow rate and duration.

12
Why develop models?

• So what about thermal models?
• Steady state heat transfer problems are solved
  using a resistance analogy.
• VTB doesnt include any thermal resistance
  models, so they needed to be created
• (new version will/does include some thermal
  models)

13
Models developed for v1.5

• Pipe flow model
• Across variable pressure (P or ?P)
• Through variable volumetric flow rate (Q)
• Pressure loss is given by
• Where f is the Moody friction factor (function of
  Re and e/D)

14
Models developed for v1.5

• That doesnt look very suitable for VTB
• Need Q Rpipe ?P - B
• Rewrite the pressure loss equation as
• f is determined using Millers expression for a
  initial estimate at the Colebrook equation

15
Models developed for v1.5

• Thermal models
• Across variable temperature (T or ?T)
• Through variable heat flow (Q)
• Heat flow is given by

16
Models developed for v1.5

• Wait! What about natural coupling?
• Natural coupling pairs are determined so that
  their product has units of power
• Fluids Pressure (N/m2) and volumetric flow
  (m3/s) ? N m/s W
• Thermal temperature (K) and heat flow (W) ? W K

17
Models developed for v1.5

• No, its okay
• Ordinarily, power is a conserved quantity (add up
  all power sources, subtract all power sinks and
  it balances to zero)
• In a thermal resistance network, the thermal
  resistive power is a conserved quantity.

18
Models created for v1.5

• Thermal Resistive Power?
• If we consider the product of temperature
  difference and heat flow as the TRP, we find it
  is conserved just as the power is in a
  traditional electrical circuit (add up all the
  TRP sources subtract all TRP sinks, and it
  balances to zero)

19
Models developed for v1.5

• Using the resistance analogy models were created
  for
• Plane conduction
• Cylindrical/Annular conduction (rods and pipe
  walls)
• Convection (assuming a convection coefficient)

20
Models developed for v1.5

• Other thermal resistance models included
• Lumped capacitance
• Thermal generation
• Intended to be a wrapper for existing models
• Model measured the across and through variables
  and computed the power dissipation
• Only valid for steady state

21
Future directions

• New base class provides substantial improvements
  for thermal and fluid modeling.
• Terminal variables can be passed as vectors,
  allowing all the relevant variables to be passed
  through one node
• Data coupling reduces the likelihood of property
  mismatch

22
Future directions

• What fluid models are under development?
• Integrated thermal pipe
• Heat Exchanger
• Pump with pump curve

23
Future directions

• What thermal models are under development?
• Constriction resistance
• Generic thermal capacitance
• Integrated heat source within an electronic
  component

24
Future directions

• Anything else? (more long term)
• Hope to begin developing plug-ins so that results
  can be animated in VXE.
• Convection models that use correlations to
  evaluate the heat transfer coefficient.