# Thermal evaluation: k-value, R-value and the energy code

Workbook: `docs/estimate/thermal-r-value.xlsx` (built by `scripts/build-thermal.py`, 12 Sep 2026; every number below is a cell on the tab named in brackets, and every tab recomputes from the Inputs tab). Assembly from `public/config/wall.config.json` and `docs/wall-spec-v1.md`. Companion documents: `standards-register.md`, `pe-package.md`, `test-matrix.md` in this folder.

The short version, in two halves. With Ken's measured R-2.1 per inch applied to all the cellular concrete (the workbook default since 12 Sep 2026): the 11.5" wall is about R-18 with the studs counted (R-25 through the concrete between studs), the house roof about R-20, the shower-house floor about R-12; the wall passes the zone 6 mass-wall U with no added insulation, the roof still needs about 3.6" of polyiso (5.7" of EPS) above the deck, and the shower house passes on the wall and needs about 2" on the roof and 1.3" under the floor. Before the test, on literature values, the wall was about R-8, the roof R-9, the floor R-5, everything failed, and the wall needed about 1.5" of polyiso. The measured number is a big improvement but its density and moisture state were not recorded, so nothing goes on a proposal until the test is repeated to ASTM C518 with the specimen described.

## 1. Ken's measured value (12 Sep 2026)

Ken's current test on the cellular concrete gives R-2.1 per inch, so k = 1/2.1 = 0.476 Btu·in/(h·ft²·°F) (0.069 W/m·K) [Inputs, measured block]. The workbook now has three yellow selector cells: the measured R per inch, "k source" (1 = measured, default; 2 = literature) and "apply the measured k to" (1 = all cellular concrete, default, because Ken tested "the cellular", one mix; 2 = the cavity fill only). Every layer R on the Wall, Roof, Floor, Code and Options tabs follows the selectors by formula; the literature values stay as the fallback and the comparison. The measured value is used as measured: the 20 percent moisture allowance applies to the literature values only, since the specimen's moisture state is whatever it had on the day.

Results with the measured k (method 3, average of the two bridging methods) [Wall, Roof, Floor, Code, Options tabs]:

| Element | k applied to ALL cellular concrete: R effective (U) | Zone 6 result | Polyiso still needed | k applied to the FILL ONLY: R effective (U) | Zone 6 result | Polyiso still needed |
|---|---|---|---|---|---|---|
| House wall, mass-wall path (U 0.060) | 17.8 (0.056); clear wall 25.0 | PASS | 0 | 9.2 (0.109); clear wall 15.6 | FAIL, gap R-7.5 | 1.2" |
| House wall, frame-wall path (U 0.045) | 17.8 (0.056) | FAIL, gap R-4.4 | 0.7" | 9.2 (0.109) | FAIL, gap R-13.0 | 2.2" |
| House roof (U 0.024) | 20.1 (0.050) | FAIL, gap R-21.5 | 3.6" (5.7" EPS) | 8.7 (0.115), unchanged | FAIL, gap R-33.0 | 5.5" |
| Shower-house wall, commercial mass wall (U 0.090) | 17.8 (0.056) | PASS | 0 | 9.2 (0.109) | FAIL, gap R-1.9 | 0.3" |
| Shower-house roof, commercial IEAD (U 0.032) | 18.9 (0.053) | FAIL, gap R-12.4 | 2.1" | 8.0 (0.125), unchanged | FAIL, gap R-23.3 | 3.9" |
| Shower-house floor, commercial mass floor (U 0.051) | 11.6 (0.086) | FAIL, gap R-8.0 | 1.3" | 5.2 (0.192), unchanged | FAIL, gap R-14.4 | 2.4" |

The mass-wall heat capacity is density-based and does not move: 6.51 Btu/(ft²·°F) with the steel, 6.33 on the concrete alone, threshold 6 [Mass wall tab]. On the total-UA path with the measured k on everything, the house fails by only 23 Btu/h·°F with the roof as cast and needs no wall board once the roof is at its code U [Options (d)].

### Where R-2.1 per inch sits

Typical published values (ASHRAE Handbook Fundamentals material tables) against the measured number [Compare tab]:

| Material | k, Btu·in/(h·ft²·°F) | R per inch | k ÷ our measured k |
|---|---|---|---|
| Steel stud | 314 | 0.003 | 659 |
| Normal-weight concrete, 145 pcf (9 to 12; 10.5 used) | 10.5 | 0.10 | 22.1 |
| Lightweight structural concrete, 100 pcf | 3.6 | 0.28 | 7.6 |
| Gypsum board | 1.1 | 0.91 | 2.3 |
| Our cellular concrete, 35 pcf, literature (dry fit) | 1.11 | 0.90 | 2.3 |
| Our cellular concrete, 30 pcf, literature (dry fit) | 0.92 | 1.09 | 1.9 |
| **Our measured cellular concrete, Ken's test** | **0.476** | **2.10** | **1.0** |
| Softwood | 0.9 | 1.11 | 1.9 |
| AAC, 31 pcf | 0.8 | 1.25 | 1.7 |
| Cellular concrete, 20 pcf, literature (dry fit) | 0.56 | 1.78 | 1.2 |
| Fiberglass batt | 0.29 | 3.45 | 0.61 |
| EPS | 0.26 | 3.85 | 0.55 |
| Mineral wool board | 0.24 | 4.17 | 0.50 |
| XPS | 0.20 | 5.0 | 0.42 |
| Polyiso, aged | 0.16 | 6.25 | 0.34 |

Plainly: R-2.1 per inch sits between AAC (R-1.25) and fiberglass batt (R-3.4). It is 2.3 times better than the 30 pcf literature value and about half of what a foam board does. On the literature curve (the k-values tab, plotted with the measured point on the Compare tab chart) a k of 0.48 corresponds to about 17 pcf material, in the 15 to 20 pcf band, so either the tested mix is lighter than the 30 pcf in the config or the foam is doing better than the published curve; the two are not the same thing for the PE (strength, pumping) or for the mass-wall check (density). Record the density and the moisture state of the test specimen (yellow cell on Inputs), put the density into the chart cell on Compare, and repeat the test to ASTM C518 on specimens of both mixes at a stated conditioning before the number goes on a proposal.

Everything from here on is the "before the test" evaluation on literature values, kept as the comparison and as what the workbook falls back to with k source = 2.

## 2. What k means and how it becomes R

k (thermal conductivity) is how much heat gets through one inch of a material: Btu per hour, per square foot, per degree F, for a one-inch thickness. Units here are Btu·in/(h·ft²·°F); divide by 6.933 for W/(m·K).

- R per inch = 1/k. A layer's R = thickness in inches ÷ k.
- An assembly's R adds the layers plus the two air films (0.68 inside a wall, 0.17 outside in winter wind; 0.61 under a ceiling, 0.92 over a floor) [Inputs].
- U = 1/R. The code tables are written in U (maximum allowed) and in R of the insulation (minimum). U is the one that fits a concrete panel, because the R table only counts insulation products and does not count concrete.

Cellular concrete is a poor insulator compared with foam or fiberglass (about R-0.75 to R-1.1 per inch against R-3.2 to R-6) but a far better one than normal concrete (R-0.08 per inch), and it is the whole wall, so the inches add up.

## 3. The working k for 30 and 35 pcf

Nobody has measured Ken's mix yet, so the workbook carries literature values [k-values tab]:

| Oven-dry density | k dry, Btu·in/(h·ft²·°F) | k dry, W/(m·K) | Published band, dry | k working (dry + 20% moisture) | R per inch, working |
|---|---|---|---|---|---|
| 30 pcf (fill) | 0.92 | 0.133 | 0.76 to 1.18 | 1.10 | 0.91 |
| 35 pcf (skin, cladding, roof, floor) | 1.11 | 0.160 | 0.90 to 1.39 | 1.33 | 0.75 |

Where the numbers come from:

- The working curve is a power fit through the ASHRAE Handbook Fundamentals design values for "foam concretes" (40 pcf → 1.3, 80 pcf → 3.0; the 60 and 100 pcf points fall on the same curve within 3 percent) [k-values, fit anchors]. ASHRAE also lists 30 pcf → 1.1 and 20 pcf → 0.8 on its perlite/vermiculite/polystyrene aggregate concrete line, which lands in the same place.
- ACI 122R (Guide to Thermal Properties of Concrete and Masonry Systems) quotes Valore's oven-dry relation k = 0.5 e^(0.02 ρ); it gives 0.91 at 30 pcf and 1.01 at 35 pcf, a second opinion within 10 percent of the fit [k-values, column E].
- ACI 523.1R (Guide for Cast-in-Place Low-Density Cellular Concrete, 50 pcf and under) has a thermal-conductivity section and ACI 523.3R covers cellular concrete above 50 pcf. The web check on 12 Sep 2026 confirmed both guides exist and cover thermal conductivity, but their tables are behind the ACI paywall, so the workbook does not quote ACI numbers it could not see. Buy 523.1R and compare its table with the k-values tab.
- The published foamed-concrete band (Ramamurthy, Nambiar and Ranjani 2009 review; Amran, Farzadnia and Abang Ali 2015 review; Mydin and Wang 2012; a 2023 mechanical-vs-chemical foaming study; a 2024 ultra-light foamed concrete review) is wide, about ±30 percent, because k depends on the foam agent, the cement, any sand, the pore size and above all moisture. The working fit sits in the upper half of that band, which is the safe side for R.
- Moisture: cellular concrete holds water and k rises with it. The 20 percent allowance is a placeholder (yellow) [Inputs]. A C518 test at the moisture condition Ken specifies replaces both the dry k and the allowance. If the dry values hold in service, the wall gains about R-2 clear (R-1.5 effective) over the numbers below.

Honesty line: these are literature ranges. Ken's mix must be tested to ASTM C518 (heat-flow meter) or ASTM C177 (guarded hot plate) before any R-value is printed on a proposal or a permit set. Type the result into the Inputs override cells and every tab updates.

## 4. Wall, roof and floor: effective R and U

Before the test: all at the literature working k (dry + 20 percent), climate-zone independent [Wall, Roof, Floor tabs with k source = 2]. Section 1 has the same table with the measured value.

| Element | Thickness | Concrete weight | R clear (through the concrete only) | R effective (steel counted) | U effective |
|---|---|---|---|---|---|
| House wall, 600S162-54 at 16" | 11.5" | 31.7 psf (+1.5 steel) | 10.2 | 7.9 (8.8 parallel path, 7.1 isothermal planes) | 0.127 |
| House roof, 800S162-68 at 16" | 13.5" solid concrete | 39.4 psf (+4 steel) | 10.9 | 8.7 (9.7 / 7.9) | 0.115 |
| Shower-house roof, 600S162-54 | 11.5" solid concrete | 33.5 psf | 9.4 | 8.0 (8.6 / 7.4) | 0.125 |
| Shower-house floor panel | 6.5" pour, studs hang 3" | 19.0 psf (+4 steel) | 6.0 | 5.2 (5.6 / 4.8) | 0.192 |

Note on the roof: the geometry (`src/geometry/roof.js`, top pour = stack minus the 6.5" ground pour) makes the whole 13.5" stack concrete, so the roof is modeled as 2" cover + 1.5" channel zone + 8" stud zone + 2" cap, all at 35 pcf. If the top pour is ever cut back to a 2" cap over the studs with the stud zone left hollow, the roof R changes and the workbook Inputs must change with it.

### Why the steel studs matter

A steel stud conducts about 240 times better than the concrete around it (k 314 vs 1.3) [Inputs]. Every 16" there is a web that carries heat straight across the wall, and the concrete feeds it sideways through the flanges. Two hand methods bracket the effect [Wall tab]:

- Parallel path: the wall is two strips side by side, a 1.625"-wide steel strip at every stud (plus the tracks; 15.3 percent of the wall face) and the clear wall between. R-8.8. This is the higher-R answer because it lets no heat spread beyond the flange.
- Isothermal planes: each plane through the wall is one temperature, and within the stud layer the 0.054" web is in parallel with the concrete. R-7.1. This is the strict lower bound on R for a stud buried in a conductor like concrete.

The truth is between them. The workbook averages the two (method selector 3 on Inputs) and reports R-7.9 / U 0.127. The accepted hand method for metal framing is the ASHRAE zone method, and ASHRAE 90.1 Appendix A publishes correction factors for insulated steel-stud walls (a 6" stud at 16" with R-19 batts is credited about R-7 for the cavity), but those factors are for k ≈ 0.26 insulation and do not transfer to a k ≈ 1 concrete cavity. For a novel assembly a code official will accept a THERM 2-D calculation stamped by the PE, or an ASTM C1363 hot-box test of a full panel. Either replaces the selector.

Why the 2" skin and the 2" cladding help: they are continuous, no steel through them, and together with the 1.5" channel zone they put 5.5" of concrete (R-4.1) in series with every stud path, so even the straight-through-steel path is R-5.0, not R-1. Without them the steel penalty would be several times larger. The 1.5" channel rows are in that continuous zone and their steel area is 0.13 percent of the face, so they are ignored [Inputs].

## 5. Is it a mass wall, and what that changes

The 2021 IECC (R202, R402.2.5) calls a wall a mass wall when its heat capacity is at least 6 Btu/(ft²·°F). Heat capacity = Σ thickness (ft) × density × specific heat; the workbook takes cellular concrete at 0.2 Btu/(lb·°F) (yellow, ASHRAE's figure for concretes; a measured value replaces it) [Mass wall tab].

- Wall heat capacity: 6.51 with the steel, 6.33 on the concrete alone. Qualifies, with a 5 to 8 percent margin.
- At 25 pcf fill it is 6.14 (still qualifies); at 20 pcf fill it is 5.76 (does not).
- If the measured specific heat comes in at 0.18 instead of 0.2, the concrete-only number drops to 5.7 and the wall is a frame wall in the code's eyes. The specific heat is worth measuring (ASTM C1784 or a DSC run) at the same time as the C518.
- The margin, plainly [Mass wall tab, sensitivity block]: 8 percent at 30 pcf fill (6.51 against 6). The skins carry about 4.1 of the 6.5 (7" of 35 pcf concrete), the steel 0.18, and the fill 0.075 per pcf of fill density, so the fill can drop to about 23 pcf (break-even 23.2) before the wall stops being a mass wall: 35 pcf fill 6.89, 30 6.51, 25 6.14, 23.2 6.00, 20 5.76, 17 5.54. If the R-2.1 mix is the 15 to 20 pcf material the literature curve suggests and it were used for everything, the wall drops to 3.44 and is nowhere near; so keep the skins at 35 pcf and use the light mix in the cavity only.

What qualifying buys: the zone 6 target drops from the frame-wall U 0.045 (R-22 effective) to the mass-wall U 0.060 (R-16.7), or 0.057 (R-17.5) when more than half of any added insulation goes on the inside [Code tab]. Under the R-value table the concrete counts for nothing, so the U-factor path is the one to use for this panel.

## 6. Pass / fail against the 2021 IECC, zone 6

Montana adopted the 2021 IECC with amendments in 2022 (Montana DEQ Residential Energy Code Handbook). The workbook carries the model-code tables; every value is marked VERIFY against the adopted Montana text and the county's climate zone (most of Montana is zone 6; the Inputs selector also does 5 and 7). The Montana amendments the handbook lists include a 4 ACH50 tightness limit and a wall-insulation option list (R-21 or R-20+5ci or R-13+10ci or R-15ci); they must be checked before any of this is relied on.

Note on table numbers: in the 2021 IECC, Table R402.1.2 is the U-factor table and Table R402.1.3 is the insulation R-value table (the task brief had them the other way round; the workbook uses the 2021 numbering).

Zone 6 residential, method 3 (average), literature working k, before the test [Code tab with k source = 2]; section 1 has the measured-value version:

| Element | U | Required U | Result | Gap as added R |
|---|---|---|---|---|
| House wall as a frame wall | 0.127 | 0.045 | FAIL | 14.4 |
| House wall as a mass wall (exterior ci or none) | 0.127 | 0.060 | FAIL | 8.8 |
| House wall as a mass wall, insulation mostly inside | 0.127 | 0.057 | FAIL | 9.7 |
| House roof | 0.115 | 0.024 | FAIL | 33.0 |
| Shower-house roof (residential row, for reference) | 0.125 | 0.024 | FAIL | 33.7 |
| Shower-house floor (residential row, for reference) | 0.192 | 0.033 | FAIL | 25.1 |

Zones 5 and 7 [Code tab, columns B and D]: the roof and floor targets are the same or tighter (floor 0.028 in zone 7); the mass wall eases to 0.082 in zone 5 (gap R-4.3) and tightens to 0.057 in zone 7.

### The shower house is not a dwelling

A shower house at an RV park or job site is a commercial occupancy: IECC Chapter 4 [CE] (C402) or ASHRAE 90.1 applies, "all other" column. The zone 6 commercial targets carried in the workbook (VERIFY) are roof U 0.032 (R-30ci basis), mass wall U 0.090, mass floor U 0.051 [Code tab, commercial block]:

| Shower-house element | U | Required U | Result | Gap as added R |
|---|---|---|---|---|
| Wall (same 11.5" panel) as a commercial mass wall | 0.127 | 0.090 | FAIL | 3.2 |
| Roof as a commercial IEAD roof | 0.125 | 0.032 | FAIL | 23.3 |
| Floor as a commercial mass floor | 0.192 | 0.051 | FAIL | 14.4 |

Semi-heated: 90.1 has a lighter envelope column for spaces whose heating is only enough for freeze protection (capacity at or above 3.4 Btu/h per ft² but under the conditioned-space threshold of 90.1 Table 3.1, on the order of 15 Btu/h per ft² in zone 6). A shower house with one small wall heater sized for freeze protection may fit that column, and its requirements are roughly half the conditioned ones; a shower house heated for comfort does not. That is a code-official call and the exact semi-heated figures must come from the adopted edition, so the workbook does not print them.

## 7. Options, ranked by what Ken would actually do

Inches are from the Options tab on the literature values (before the test); polyiso at R-6 per inch aged, EPS at R-3.8, XPS at R-5, mineral wool board at R-4.2 (all yellow, product data governs). Round up to stock thickness. With the measured value the ranking holds but the wall drops out of item 2 (it passes the mass-wall U on its own) and the roof board shrinks to about 3.6" of polyiso; see section 1.

1. **Insulate the roof above the deck, under the membrane.** This is not optional under any path: the roof is 33 R short of the residential ceiling U and 23 R short of the commercial roof U. House roof: 5.5" polyiso (8.7" EPS, 6.6" XPS) [Options (a)], about +1 psf, and the membrane goes over the board (a standard low-slope roof detail: tapered polyiso is also how the flat shower-house roof would get its drainage slope instead of a thicker concrete build-up). Shower-house roof, commercial: 3.9" polyiso (6.1" EPS). Nothing on the wall side matters until this is done, because in the total-UA trade the roof deficit alone is bigger than the whole reference budget [Options (d), first block].

2. **Take the mass-wall path and add about 1.5" of polyiso to the wall** (1.47" by the series estimate; recomputed with the board in the assembly the wall lands at U 0.059 against 0.060, PASS) [Options (a) and the check block]. Alternatives: 2.3" EPS or 1.8" XPS or 2.1" mineral wool. Placement:
   - Outside, over the cladding, under a finish: wall 13.0" with polyiso, no interior change, and the exterior-insulation target (0.060) applies. It needs a finish (the walls are natural grey concrete by default), so it changes the product.
   - Between the outside stud flanges and the cladding, cast in: the cleanest thermal break and no visible change, but the cladding then hangs on fasteners through the board (PE item) and the board must survive the pour.
   - Inside, between the skin and a furred drywall finish: the interior-insulation target (0.057) applies, the gap grows to R-9.7 (1.6" polyiso), and the cast finish face, which is the point of the panel, gets covered.
   Weight added is under 0.3 psf in every case [Options (a)].

3. **Use the total-UA alternative (R402.1.5, what REScheck computes) or the R405 simulated-performance path to trim the wall board.** With the roof at its code U and U-0.27 windows the wall needs U 0.066 (R-15.1), 1.2" polyiso instead of 1.5"; with U-0.22 triple-pane windows, 0.86" [Options (d), continued block]. No window on the market carries the wall with zero board (the required window U goes negative). R405 (annual simulation by a HERS rater) adds a modest thermal-mass credit on top and is the path most likely to carry a concrete house with the least board; it costs a rater and a model.

4. **Deeper studs (8" 800S162 in the wall, 13.5" wall).** Gains about R-1 effective (R-1.8 clear wall; the deeper steel takes some back), adds 5 psf and 2" to every jamb and sill, and still leaves an R-7.8 gap [Options (c)]. It does not close a zone-6 gap on its own; it is a structural decision, not a thermal one.

5. **Lighter fill (25 or 20 pcf).** Buys about R-0.4 at 25 pcf and R-0.9 at 20 pcf effective, because the fill is only 4.5" of the 11.5" and the k curve flattens at low density; at 20 pcf the wall stops being a mass wall (heat capacity 5.76) and the gap under the frame-wall rules becomes larger, not smaller [Options (b), Mass wall tab]. A 20 to 25 pcf cellular concrete also has roughly a quarter to a half of the 30 pcf strength, is harder to pump without collapse, and shrinks more; the fill braces the studs and holds the cladding screws, so any change goes through the PE package (`pe-package.md`). Not recommended for thermal reasons.

6. **(e) Unfilled cavity with rock wool, cellular cladding on.** The 5" skin pour (2" skin + 1.5" channel zone + 1.5" of the studs, cellular at the skin k in use), then 4.5" of mineral-wool batt in the stud cavity instead of the pumped fill, then the 2" cellular cladding; films 0.68 / 0.17 [Options (e)]. The cavity is credited by the ASHRAE 90.1 Appendix A steel-stud table (Table A9.2B, Table A3.3 in older editions): 6" studs at 16" o.c. with R-19 → correction factor 0.37, effective R-7.0 (both yellow). That factor replaces the parallel-path / isothermal-planes averaging for a batt cavity because that table is what a code official uses for insulated steel-stud walls. The wall is then 7" of cellular concrete plus the corrected cavity; it is no longer a mass wall (heat capacity 4.26 against 6, skins plus steel only), so it is judged as a steel-framed wall: residential frame-wall U 0.045 in zone 6, and for the shower house the commercial steel-framed wall U (0.064 carried from IECC C402.1.4, VERIFY; 90.1 is tighter, about 0.049 in zone 6).

   | k case | Filled wall R eff (U), mass-wall target | (e) wool wall R eff (U) | (e) vs residential frame U 0.045 | (e) vs commercial steel-framed U 0.064 |
   |---|---|---|---|---|
   | Measured R-2.1/in on all cellular | 17.8 (0.056), PASS | 22.6 (0.044) | PASS, no board | PASS |
   | Measured on the fill only, literature skins | 9.2 (0.109), 1.2" polyiso | 13.2 (0.076) | FAIL, 1.5" polyiso | FAIL, 0.4" |
   | Literature, before the test | 7.9 (0.127), 1.5" polyiso | 13.2 (0.076) | FAIL, 1.5" polyiso | FAIL, 0.4" |

   The wool wall beats the filled wall thermally in every case (the fill is R-2.1 per inch at best, the corrected batt is R-1.6 per inch, but the filled wall also pays the full steel penalty), and with measured skins it passes the frame-wall U on its own. What changes besides thermal:
   - Loses the mass-wall status (4.3 against 6) and the thermal mass itself, the flywheel that flattens the daily swing and earns credit in an R405 simulation; the target moves from 0.060 to 0.045.
   - Loses the composite action and bracing of a full cavity (the fill braces the studs, stiffens the panel and carries the cladding screws), so the PE package must re-check the stud design and the cladding attachment; and it loses the sound and fire benefit of a solid wall.
   - Needs a warm-side vapor retarder over the skin pour and a dew-point check on the cladding: a 2" concrete cladding on the cold side of a batt cavity sits below the dew point of indoor air for much of a Montana winter and the batt does not stop vapor. That is an energy consultant / WUFI hygrothermal check, not a hand calculation. Mineral wool tolerates wetting better than fiberglass, which is why it is the batt named.
   - Gains: no pumping on site, an open cavity for the trades until the cladding goes on, and a lighter wall (about 22 psf against 33).
   - Cost is not priced here: the estimate workbook's Walls tab carries the cavity-fill line (yd³ at the mix price plus pump labor); the batt, the vapor retarder and any extra fasteners replace it there.

What the expectation was and what the numbers say: the brief expected the wall around R-10 to R-14 effective and the roof similar. On literature values, with the moisture allowance and the steel averaged, the wall lands at R-7.9 (R-8.8 parallel path; R-10.2 clear; about R-12 clear if the dry k holds) and the house roof at R-8.7. Zone 6 therefore needs the mass-wall path plus about 1.5" of continuous insulation on the wall, or the performance path with about 1" plus good windows, and the roof needs real added insulation in every case. That is a couple of R points below the low end of the expectation, most of which is the 20 percent moisture allowance and the choice to average the two bridging methods. Ken's measured R-2.1 per inch moves the wall to R-17.8 and the roof to R-20.1 (section 1), above the expectation; the roof still needs board, the wall does not.

## 8. Test plan

1. **ASTM C518 (or C177) on the mix**, both densities (35 pcf skin/cladding/roof mix and 30 pcf fill), oven-dry and at an in-service moisture condition (ACI 122R gives the conditioning conventions). Two or three specimens each, with the density, age and moisture state of every specimen written on the report; that is what Ken's current R-2.1 test is missing. This replaces the yellow k cells and the moisture allowance. Cost is a few hundred dollars per specimen at a commercial lab; it is the cheapest item on this list and the one that makes every other number real.
2. **Specific heat** (ASTM C1784 or a DSC run) on the same specimens, because the mass-wall status rides on a 5 to 8 percent margin at the assumed 0.2.
3. **Density and moisture** of cores from a cured panel, so the C518 conditioning matches what a wall actually holds after a Montana winter.
4. **THERM 2-D model** of the stud-in-concrete section by the PE (or an energy consultant) using the C518 k: this is the calculation a code official will normally accept in place of the hand average, and it decides whether the wall is R-8 or R-9.
5. **ASTM C1363 hot-box test on a full 4' × 8' or 4' × 10' panel** only if the mass-wall or performance path is contested by the official, or when the panel goes to market with a published R-value. It is the expensive one (a certified lab, a panel shipped, several thousand dollars) and it should wait until items 1 to 4 have set the design.
6. **Code check with the adopted Montana text**: confirm the county zone, the Montana wall amendment, the 4 ACH50 limit, the slab-edge R-10, and whether the local jurisdiction accepts REScheck or requires R405 for a mass wall.

## 9. Where the assumptions are (all yellow in the workbook)

- Ken's measured R-2.1 per inch: density and moisture state of the specimen unknown (yellow "record it" cell); applied to all cellular concrete by default (k source = 1, apply = 1), used as measured with no moisture allowance [Inputs].
- Chart density for the measured point: 30 pcf assumed, replace with the test density [Compare].
- k for both densities when k source = 2: literature fit, not a test [Inputs, k-values].
- 20 percent moisture allowance on the literature k only [Inputs].
- Specific heat 0.2 and 1.5 psf of wall steel for the heat-capacity check [Inputs].
- Method selector 3 (average of parallel path and isothermal planes) [Inputs].
- Steel design thickness 0.0538" (54 mil) and 0.0713" (68 mil) [Inputs].
- Roof modeled as solid concrete over the full stack per the geometry [Inputs, roof block].
- Floor: the 3" of stud below the pour treated as part of the outside surface [Floor tab].
- Every code number: 2021 IECC model code, VERIFY against the adopted Montana code and the county zone [Code].
- Insulation R per inch and weights; window and door U-factors for the UA trade [Inputs].