as QCoil is negative in cooling mode, directly using
m_air = Q_coil / (T_coil_out - T_coil_in) / 1005.0
will result in negative air flow, as well as downstream fan power

Here QCoil should be abs(QCoil)

…ower

per Mike's comment,
#10649 (comment),
Q_coil should not be negative, remove the abs(Q_coil) fix

Q_evap_req will be negative if Q_cond_req <
CompEvaporatingPWRSpd(CounterCompSpdTemp). This leads to negative compressor
speed. Temtatively fix it like this.

According to comment here
#10649 (comment),
The h_comp_out and h_comp_out_new could keep getting farther and farther from
each other if h_cond_IU_in can only increase in the iteration. Addressing this
by adding a direction where h_cond_IU can decrease as well.

…ower

…ower

This reverts commit 35d4b69.

When Q_cond_req < ComoEvaporatingPWRSpd(1), Q_evap_req will be negative, the
calculated compressor speed will be negative. The case where Q_cond_req <
ComoEvaporatingPWRSpd(1) need to be considered, where compressor speed is
dependent on the compressour power curve output value.

CompSpdActual argument VRFOU_CompCap is an int, but the value passed in was
originally a real. This casting from int to real will truncate the value. If the
value is less than 1, it will be set to 0.

…ower

…ower

C_cap_operation was used in the evaporative load adjsutment but not power. not
sure if this is intended. Maybe just remove this for now

Q_evap_req is no longer used in the calculation, remove it

how r is derived:

in the original code
Q_evap_req = Q_cond_req - CompEvaporatingPWRSpd(CounterCompSpdTemp);
I think the intent was
Q_cond_req = Q_req = Q_evap_req + compressor heat
but compressor heat might not actually be CompEvaporatingPWRSpd(CounterCompSpdTemp)

let
k be the speed level
C be the short form for C_cap_operation
PWR(.) be the short form of CompEvaporatingPWRSpd(.)
CAP(.) be the short form of CompEvaporatingCAPSpd(.)
spd(.) be the short form of this->CompressorSpeed(.)
deltaPWR be PWR(k) - PWR(k - 1)
deltaCAP be CAP(k) - CAP(k - 1)

I think compressor heat should be this instead of just PWR(k)
compressor heat = deltaPWR * r + PWR(k - 1)
so
Q_cond_req = Q_req = Q_evap_req + deltaPWR * r + PWR(k - 1) <---- eq 1

we also know this from the CompSpdActual calculation equation that
CompSpdActual = Spd(k - 1) + deltaSpd / deltaCAP * (Q_evap_req * C - CAP(k - 1))
so we call the following r
r = (Q_evap_req * C - CAP(k - 1)) / deltaCAP
so
CompSpdActual = Spd(k - 1) + deltaSpd * r

arranging terms in eq 1
Q_cond_req - deltaPWR * r - PWR(k - 1) = Q_evap_req
(Q_cond_req - deltaPWR * r - PWR(k - 1)) * C - CAP(k - 1) = Q_evap_req * C - CAP(k - 1)
((Q_cond_req - deltaPWR * r - PWR(k - 1)) * C - CAP(k - 1)) / deltaCAP = (Q_evap_req * C - CAP(k - 1)) / deltaCAP
((Q_cond_req - deltaPWR * r - PWR(k - 1)) * C - CAP(k - 1)) / deltaCAP = r
(Q_cond_req - deltaPWR * r - PWR(k - 1)) * C - CAP(k - 1) = deltaCAP * r
(Q_cond_req - PWR(k - 1)) * C - CAP(k - 1) = deltaCAP * r + deltaPWR * r * C
so
r = ((Q_cond_req - PWR(k - 1)) * C - CAP(k - 1)) / (deltaCAP + deltaPWR * C)
  = ((Q_cond_req - PWR(k - 1)) - CAP(k - 1)/C) / (deltaCAP/C + deltaPWR)

in the special case where k = 1, then k - 1, PWR(k - 1) and CAP(k - 1) will all be 0

r = ((Q_cond_req) * C) / (CAP(1) + PWR(1) * C)
  = (Q_cond_req) / (CAP(1)/C + PWR(1))