Material testing update

In continuation on the material preparation and its assessment we prepared the first workshop test matrix for the Phase II TPU envelope material.

This update focuses on the next step: what we are testing, why the current targets changed, what the active plan is, and the first recorded outputs from the strip-loop tests.

What we are testing

The current tests are focused on the delivered `0.13 mm` aliphatic polyether TPU film and the heat-welded lap seam. The immediate questions are:

• how much the material stretches under load
• whether the weld holds under the intended proof-of-concept loads
• whether the strip keeps creeping during the hold
• how much length recovers after unloading
• whether longitudinal and transverse roll directions behave differently
• whether the results are good enough to move toward a closed pressure article

The directionality question matters. The supplier reference sheet lists different tensile strengths:

Direction	Tensile strength
Longitudinal	576 N / 5 cm
Transverse	459 N / 5 cm

That may affect the eventual cutting plan, because the weaker direction could become the governing case if it carries the main hoop load.

Revised proof-of-concept pressure basis

Earlier design work used a conservative `300-500 Pa` operating frame, with higher pressures kept as proof or transient references.

For the current Phase II-A proof-of-concept, the target is softer:

• expected maximum speed: about `37 km/h`
• intended demo/operation duration: about `2 h`
• possible inflated duration: up to about `12 h`

Current working interpretation:

Full-envelope pressure	Interpretation
250 Pa	nominal PoC pressure
300 Pa	strong PoC operating pressure
350 Pa	upper operating / gust margin
500 Pa	proof test, not normal operation

The purpose of the strip tests is not to force the material to meet the old continuous `500 Pa` assumption. It is to find the practical creep, recovery, and weld-performance limits, then use those results to decide the next pressure article.

Strip-loop setup

The restarted test set uses 12 welded loops:

• strip width: `100 mm`
• strip length: `400 mm`
• lap seam: `50 mm`
• weld temperature: about `220 C`
• loops `1-4`: longitudinal / roll direction
• loops `5-12`: transverse / cross-roll direction
• average ambient temperature 20C

The current test mass is `6.655 kg`, equivalent to about `261 Pa` full-envelope pressure in this strip-loop setup.

Test log

Loop #1: Longitudinal / roll direction

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-07 16:10	0h	6.655 kg	~261 Pa	420 mm	immediate stretch from `400 mm`; sample holding
2026-06-07 19:10	3h	6.655 kg	~261 Pa	440 mm	holding; seam/sample visually happy
2026-06-08 06:30	14 h 20 min	6.655 kg	~261 Pa	460 mm	survived overnight; no visible seam
after unload	instant	none	n/a	420 mm	immediate recovery after removal from jig
2026-06-08 18:10	11 h 40 min	none	n/a	400 mm	recovered to original length

Notes:

• original unloaded length before loading: `400 mm`
• the exact `12 h` mark was `2026-06-08 04:10`
• the practical morning check was taken at `14 h 20 min`
• after nearly 12 h off-load, the sample had recovered back to the original 400 mm

Loop #2: Longitudinal Middle-Margin Hold

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-09 20:15	0 h	7.700 kg	~302 Pa	440 mm
2026-06-09 22:15	2 h	7.700 kg	~302 Pa	470 mm
2026-06-09 22:15	instant	none	n/a	420 mm	Instantly recovered after removal from jig
2026-06-10 8:15	12h	none	n/a	400mm	Completely recovered

Loop #3: Longitudinal Upper-Margin Hold

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-09 06:30	0h	9.085 kg	~356 Pa	460 mm	Immediate elongation from 400 mm; 30-60 min upper-margin test started
2026-06-09 07:10	40 min	9.085 kg	~356 Pa	480 mm	Removed from jig; seams ok-ish, with minor edge peel-away observed
2026-06-09 07:10	instant	none	n/a	415 mm	Instantly recovered after removal from jig
2026-06-09 18:30	11 h 20 m	none	n/a	400	Recovered completely

Notes:

• original unloaded length before loading: `400 mm`
• loop 3 is longitudinal / roll direction
• this load is slightly above the nominal 350 Pa upper operating / gust-margin target
• minor seam-edge peel on loop 3, after similar observations on loops 5 and 6, supports adding a general seam inspection and repeat-load/cycling check after the current set

Loop #5: Transverse / cross-roll direction

Loop `5` is important because transverse behaviour may be the governing design case.

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-08 06:30	0h	6.655 kg	~261 Pa	420 mm	immediate stretch from `400 mm`; second transverse sample started
2026-06-08 18:10	11 h 40 min	6.655 kg	~261 Pa	~460-470	weld had no apparent damage
2026-06-08 18:10	instant	none	n/a	430 mm	immediate recovery after removal from jig
2026-06-09 06:20	12 h 10 min	none	n/a	~410 mm	recovered further; similar seam-edge peel defect observed as loop 6

Notes:

• original unloaded length before loading: 400 mm
• loop #5 is the first governing-direction transverse check at the nominal PoC load.
• loaded length before removal was estimated by feel/handling rather than recorded as a precise ruler measurement.
• the observed seam-edge peel on loops 5 and 6 suggests the weld process needs inspection before treating transverse nominal-load survival as a clean pass.

Loop #6: Transverse / cross-roll direction

Loop `6` is to confirm observations through the loop #5.

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-08 18:10	0h	6.655 kg	~261 Pa	430 mm	immediate stretch from `400 mm`; transverse sample started
2026-06-08 18:10	11 h 40 min	6.655 kg	~261 Pa	~460-470	weld had no apparent damage
2026-06-08 18:10	instant	none	n/a	430 mm	immediate recovery after removal from jig
2026-06-09 16:25	22h	none	n/a	410 mm	recovery/set checked; this may include cutting, welding, or measurement error

Loop #9: Transverse / cross-roll direction

Loop `9` is important because it was first short higher-load samples.

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-09 15:40	0h	9.095 kg	~356 Pa	460 mm
2026-06-09 16:25	45 min	9.095 kg	~356 Pa	540 mm
2026-06-09 16:25	instant	none	n/a	420 mm	immediate recovery after removal from jig
2026-06-10 6:25	14 h	none	n/a	400 mm	Complete recovery

Loop #10: Transverse / cross-roll direction

Loop `10` is loop `9` cross-check

Time	Elapsed	Load	Equivalent	Length	Observation
2026-06-09 16:25	0h	9.095 kg	~356 Pa	460 mm
2026-06-09 17:05	40 min	9.095 kg	~356 Pa	540 mm
2026-06-09 17:05	instant	none	n/a	420 mm	immediate recovery after removal from jig
2026-06-09 22:15	5 h 10 min	none	n/a	400	complete recovery, weld clean

Follow-up `30 mm` seam check

Because the 50 mm overlap samples often appeared to peel at the outer edge while retaining a strong central bond, we made three fresh longitudinal loops with a deliberate 30 mm weld overlap at about 220 C.

All three were tested at 7.7 kg, about 302 Pa equivalent:

Loop	Test history	Result
#13	2 h, then another 1 h 50 min repeat cycle	no damage observed
#14	about 4 h 25 min hold repeated cycle	no damage observed
#15	about 1 h 20 min, then another 2 h 10 min repeat cycle	one external-corner defect was marked; later judged likely setup damage, and it did not propagate under the repeat cycle

This is a useful result because it separates the apparent material/weld-center strength from the edge-peel problem. A deliberate `30 mm` seam, aligned under the effective hot/pressure zone, appears stable at the strong-operating loop load.

Final results

The first controlled longitudinal result is encouraging: the 220 C / 50 mm lap seam survived overnight at about the nominal PoC pressure equivalent without visible seam damage, and recovered to the original 400 mm length after nearly 12 h off-load.

The nominal transverse samples also survived their holds, but loops 5 and 6 both showed seam-edge peel. That changes the interpretation: the material strength is still promising, but the weld process needs closer inspection before we treat the current seam as qualified.

The higher-load samples are useful but mixed. Loops 3, 9, and 10 are encouraging because they recovered to about the original 400 mm length after unloading, even after upper-margin loading. At the same time, the loaded elongation at the upper-margin load is large, and several samples now point back to seam-edge fusion as the issue to control.

The ~300 Pa / 2 h strong-operating checks have now run in both directions. Loop 2 covered the longitudinal case and recovered to about 400-410 mm. Loops 7 and 8 repeated the same load in the transverse direction; both reached about 500 mm while loaded and later recovered to 400 mm. That is encouraging enough to keep the strong-operating case alive, while still watching seam quality closely.

The proof-zone tests near 12.5 kg / ~490 Pa are a different story. Loop 4 in the longitudinal direction reached about 620 mm after 10 min and instantly recovered to 420 mm. Transverse loops 11 and 12 stretched much more, reaching about 920 mm and 1020 mm, with instant recovery only to 460 mm and 520 mm. Unless later recovery is surprisingly complete, this looks like the beginning of a practical no-go zone for the current loop geometry and weld process.

The post-test seam review also shows a pattern. The maximum peel-away depth was usually worse on the outer seam edge than the inner seam edge, while the center of the seam often still appeared strongly bonded. That points toward a weld-process issue more than a simple material-strength issue: alignment, iron heat distribution, pressure distribution, backing compliance, the baking-paper stack, or the thermal sock may be giving a reliable central bond but weaker edges.

The follow-up 30 mm seam loops support that interpretation. When the weld was deliberately centered on the apparent effective bond zone, the loops survived repeated strong-operating load cycles, and the one known damaged corner did not grow.

The main caution is that survival is not the same as qualification. The real decision points are:

• creep rate during the hold
• permanent set after unloading
• seam-edge peel or weld-edge movement
• repeatability across more samples
• comparison between longitudinal and transverse directions

The strip-loop gate is now good enough to move toward an air-only pressure article, with the weld process changed from a nominal 50 mm overlap to a controlled 30 mm effective seam. The next question is no longer whether coupons can survive the strong-operating load; it is whether a closed article can be built leak-tight, hold pressure, and keep its seams stable.

The 400 mm x 1200 mm cylinder is the cleaner next article for validating the real hoop/longitudinal seam geometry, unless a smaller pillow is needed first to rehearse fittings or gauge plumbing.

Finally huge thanks to Serge for preparing awesome roll-holder jig, these rolls are seriously heavy! Thank you Serge! 🙂