ASHRAE Journal - June 2020 - 58

COLUMN DIGITAL LABORATORY

FIGURE 3 Temperature distribution in the underfloor plenum: (a) with thermal
radiation calculation; (b) without thermal radiation calculation.

TABLE 3 Underfloor office example-averaged results.
VARIABLE

UNITS

WITH RADIATION MODEL

WITHOUT RADIATION
MODEL

T42"

°F (°C)

73.9 (23.3)

80.6 (27)

DThf

R (K)

2.1 (1.1)

7.8 (4.3)

Twindow

°F (°C)

77.3 (25.1)

118.9 (48.3)

DTdecay

R (K)

8.3 (4.6)

10 (5.5)

Qslab

Btu/h (W)

9,513 (2788)

19,329 (5665)

Qoffice

Btu/h (W)

12,942 (3793)

3,108 (911)

Qplenum

Btu/h (W)

22,454 (6581)

22,437 (6576)

is exactly the same; it's the way the heat gain is distributed (radiant vs. convective) that differs between the
two simulations.
In the simulation without a thermal radiation model,
the higher thermal stratification is due to the absence
of radiant heat transfer from a warm ceiling to a cool
floor, which has a significant impact on the indoor environment in this space. One last item to highlight in the
office portion of the simulation is the temperature of
the window (interior surface of the glass) which is 42°F
(23.3°C) higher in the simulation without thermal radiation. This is because the naturally radiant solar load is
implemented as purely convective when the thermal
radiation model is not active.
Although this has the effect of "capturing" that solar
load at the window, the uncharacteristically high surface temperature implies that this effect is unrealistic.
One could argue that the solar load can be manually
split up between the window and the surfaces it would
reach (in the absence of a radiation model), which
would be closer to reality. However, this approach
would require a lot of assumptions to be made as far as
what portion of that load is applied to which surfaces,
thereby turning it into a guessing game. It is also worth
remembering at this point that the underfloor system
in this office has grilles at the perimeter that "wash"
the glass with cold, high-velocity air, yet the radiant
effect is still significant.
Shifting the focus to the conditions in the underfloor
plenum, the results depicted in Figures 3a and 3b surprisingly show very little difference in the temperature
distribution. Even though the average thermal decay
(Table 3) is slightly higher in the case with no thermal
58

ASHRAE JOURNAL

ashrae.org

J U N E 2020

A. With Thermal Radiation Calculation

B. Without Thermal Radiation Calculation

radiation, the air in the plenum is equally well mixed in
both cases. So, what's going on here? The go-to answer
is that the air speeds in the plenum are around 500 fpm
(2.5 m/s) and therefore much higher than in the occupied space; as such, the radiant heat transfer should be
small.
While that explanation is technically correct, the
details are far more interesting. If we examine the heat
transfer into the underfloor plenum (Table 3) from the
floor below and the office space above (calculated by the
simulation, not assumed), we see that the sum of the two
is the same for both cases, but there is more heat gain
from the floor below than from the office in the case
without thermal radiation. Since the total heat gain and
the supply airflow are the same in both simulations, the
temperatures should be the same (on average), which is
what the results show. Without thermal radiation heat
https://www.ashrae.org/
ASHRAE Journal - June 2020

Table of Contents for the Digital Edition of ASHRAE Journal - June 2020
