Materials and processes

Why scale changes the materials questions

Most PocketQube teams build with commercial terrestrial parts and inherit their materials rules from three sources: the CubeSat Design Specification for rails, coatings and outgassing, the ECSS standards that the PocketQube Standard cites, and passive-magnetic design rules developed for 1U–3U CubeSats[1]. Each was written for a different size. Area, volume, inertia, heat capacity and magnetic near fields scale with different powers of length, so a rule that is conservative for a CubeSat can be inadequate, irrelevant or needlessly strict for a PocketQube.

Between a 1P and a 1U, geometric similarity divides mass and volume by 8, area and body-mounted power by 4 and the moment of inertia by 32. The standards do not keep similarity: at their mass limits a 1P has 4.7% of the inertia of a 1.33 kg 1U, its areal density is 0.75 times as large, and the field of a magnet of given moment is eight times stronger at its inner walls[1].

The clauses that matter

The PocketQube Standard, Issue 1, contains a short list of requirements with direct materials consequences[2]:

The CubeSat Design Specification lists 7075, 6061, 6082, 5005 and 5052 for CubeSat structures and asks for hard-anodized rails[3].

Alloys and stress corrosion

Material	Density (g/cm³)	E (GPa)	Yield (MPa)	Tensile (MPa)	α (ppm/K)	k (W/(m·K))	SCC resistance (ECSS-Q-ST-70-36C)	PocketQube clause
Al 6061-T6	2.70	68.9	276	310	23.6	167	High resistance	Mat-02
Al 6082-T6	2.70	70.0	260	310	23.4	172	High resistance	Mat-02
Al 7075-T6	2.81	71.7	503	572	23.4	130	Low resistance	Mat-02
Al 7075-T73	2.81	72.0	435	505	23.4	155	High resistance	Mat-02
Al 2024-T3	2.78	73.1	345	483	23.2	121	Low resistance	Mat-04
AlSi10Mg, LPBF as built	2.67	≈70.0	270	461	—	—	Not listed	Mat-04
Mg AZ31B-H24	1.77	45.0	220	290	26	96	Not listed	Gen-03, Mat-04
Ti-6Al-4V, annealed	4.43	113.8	880	950	8.6	6.70	High resistance	Mat-04
17-7PH, TH1050	7.64	204.0	1,170	1,310	11	16	By heat treatment	Mat-04
FR-4, in plane	1.85	≈22.0	—	300	14–17	0.30	—	Mat-02
Windform XT 2.0, SLS	1.10	8.93	—	83.8	—	—	—	Mat-04
PEEK, unfilled	1.30	≈3.80	98	—	47	0.25	—	Mat-04
ULTEM 9085, FDM (XZ)	1.34	2.15	47	69.0	65	—	—	Mat-04

Materials review, Table 2. Mat-02: recommended; Mat-04: launch-provider consultation; Gen-03: flammable materials. FR-4 expands 50–70 ppm/K through its thickness. Design allowables must come from qualified sources.

In 6xxx (Al–Mg–Si) alloys, peak strength comes from needle-shaped β″ precipitates; the alloys combine moderate strength with good machinability, an even anodizing response and high stress-corrosion resistance, and ECSS-Q-ST-70-36C places the whole 6000 series, in all tempers, in its high-resistance table[4]. In 7xxx (Al–Zn–Mg–Cu) alloys the peak-aged T6 temper reaches about 500 MPa yield strength but is prone to intergranular stress-corrosion cracking, especially under short-transverse tension[5]. Over-ageing to T73 or T7351 restores the resistance at a strength penalty of roughly 10–15%.

Two practical conclusions follow[1]. A 7075-T6 backplate, the most available temper of an alloy that PQ-Mat-02 recommends, formally needs a customer-approved stress-corrosion evaluation form under the very standard the clause invokes, whereas a T73 plate does not. And because the backplate is not strength-critical, the strength of 7075 buys nothing there: 6061-T6 or 6082-T6 do the job with high stress-corrosion resistance and a more uniform hard-anodic film.

The backplate is interface-critical

Under an illustrative quasi-static load of 20 g, a 250 g unit presses 49 N on the 256 mm² of clamped backplate edges: a mean bearing pressure of 0.19 MPa, more than two orders of magnitude below the strength of any candidate material[1]. The plate is governed by flatness, thickness and width tolerance, edge condition, surface hardness and the tribology of the clamped strips. With its thickness fixed, its mass depends only on density:

Material	Mass (g)	Share of 250 g	Flexural rigidity D (N·m)	D ÷ D of 6061-T6	Stiffness index E^(1/3)/ρ
FR-4, in plane	11.0	4.4%	7.6	0.29	1.51
Mg AZ31B-H24	10.5	4.2%	17.5	0.66	2.01
Al 6061-T6	16.0	6.4%	26.4	1.00	1.52
Al 7075-T73	16.7	6.7%	27.6	1.04	1.48
Ti-6Al-4V, annealed	26.3	10.5%	44.0	1.67	1.09
17-7PH, TH1050	45.4	18.2%	75.6	2.86	0.77

D = E t³ / [12 (1 − ν²)] with ν = 0.12 (FR-4), 0.35 (AZ31B), 0.33 (aluminium), 0.342 (Ti-6Al-4V) and 0.28 (17-7PH). Magnesium falls under the flammability clause PQ-Gen-03.

For panels of free thickness the mass-optimal material indices are $E^{1/3}/\rho$ for stiffness and ${\sigma }_f^{1/2}/\rho$ for strength[6]. On these indices magnesium and the laser-sintered composite Windform XT 2.0 lead in stiffness and FR-4 in strength, which is why boards that double as structure are mass-efficient; the price is a coupling of structural, thermal and electrical design.

Additive manufacturing

Laser powder-bed-fused AlSi10Mg reaches as-built yield strengths of 233–270 MPa, comparable with 6061-T6, and topology-optimized AM structures promise 30–60% mass savings. Qualification lags behind: ECSS-Q-ST-70-80C offers no de-rated pathway for parts below 100 g, and AA6061-T6 remains the de facto PocketQube structural alloy[7].

Cold welding and fretting

In air, oxides and adsorbed layers keep metal asperities apart. Small oscillating slides during launch vibration (fretting) or repeated impacts break these films, and in vacuum they cannot re-form quickly enough: clean asperities bond at room temperature[8]. Recent campaigns observed adhesion at about 1 Pa and below; self-mated silver, copper and 17-7PH exceeded the 12 000 mN limit of the test, and a sputtered MoS₂ coating on 17-7PH against Ti-6Al-4V failed after about 40 fretting cycles in vacuum[9]. The best-known flight consequence is the Galileo high-gain antenna, which did not fully deploy in 1991 after three ribs cold-welded once vibration had worn their dry lubricant[10].

A PocketQube deployer has every ingredient of this failure: two clamped strips per unit and contacts between stacked units, preloaded and vibrated in qualification tests and at launch, then held in vacuum for hours to weeks before ejection. Both the CubeSat and the PocketQube standards respond with hard anodizing of the aluminium contact surfaces.

Hard anodizing

Type III (hard) anodizing grows a thick, dense, porous oxide in cold sulfuric acid at high current density, commonly specified at 50 µm within 25–75 µm[11]. Its ceramic character, not an absence of pores, prevents adhesion. Five side effects matter at PocketQube tolerances[1]:

• Growth. About half the film grows outward: a 50 µm film adds about 50 µm to the 58.0 mm backplate width, half of the ±0.1 mm tolerance band. Machine undersize and measure the film after coating.
• Alloy response. Copper-rich 2xxx alloys and high-silicon cast or AM alloys form thinner, softer films; pulsed-current hard anodizing of AM AlSi10Mg reached about 280 HV at best[12].
• Fatigue debit. Hard anodizing cut the high-cycle fatigue strength of 6082 by 50–62%, against 8–50% for conventional anodizing[13]. Acceptable for the lightly loaded backplate, not for anodized springs or hinges.
• Sealing and impregnation. Sealing improves corrosion resistance but lowers hardness; PTFE-impregnated coats reduce friction but add an organic phase that must meet the outgassing limits.
• Electrical isolation. Anodic films insulate the chassis from the deployer, which affects grounding and electrostatic-discharge design.

Recommended practice

Prefer 6061-T6 or 6082-T6 for backplates and avoid self-mated identical metals at any separable contact; specify anodizing type, thickness and sealing on the drawing; round edges and keep sliding faces below Ra 1.6 µm, as for CubeSat rails; inspect film thickness, dimensions and flatness after coating; and after vibration testing run a vacuum dwell and ejection test of a flight-representative backplate–rail pair[1].

Thermal cycling

A PocketQube has little heat capacity per square metre of radiating surface, so it swings harder in every eclipse:

Case	Minimum (°C)	Maximum (°C)	Swing (K)	Fastest cooling (K/min)
1P, 200 g	−15.0	21.8	36.8	1.50
1P, 250 g	−10.8	19.5	30.3	1.16
1U, 1.33 kg	−6.6	16.8	23.4	0.83
1U, 2.00 kg	−2.2	13.6	15.8	0.52

α = ε = 0.85, c_p = 858 J/(kg·K), albedo 0.30, Earth infrared 237 W/m², internal dissipation 0.25 W (1P) and 1.0 W (1U), β = 0°.

These are bulk values; SMOG-1’s finite-element model predicted panel extremes of about +15 and −26 °C[14]. At about 5500 eclipses a year, every interface sees tens of thousands of cycles over a mission.

Expansion mismatch and solder fatigue

Triple-junction cells grow on germanium (about 6 ppm/K), FR-4 expands by 14–17 ppm/K in plane, ceramic chip components by 6.5–7 ppm/K, copper by 17 and aluminium 6061 by 23.6. The Engelmaier model for rigid leadless joints links the swing to joint life[15]:

Thermal case	Swing (K)	Δα = 11 ppm/K	Δα = 6 ppm/K	Δα = 2 ppm/K
1U, 1.33 kg	23.4	9.9 × 10³1.8 years	4.6 × 10⁴8.3 years	7.3 × 10⁵132 years
1P, 250 g	30.3	5.1 × 10³0.9 years	2.4 × 10⁴4.3 years	3.8 × 10⁵68 years
1P, 200 g	36.8	3.1 × 10³0.6 years	1.5 × 10⁴2.6 years	2.3 × 10⁵42 years

Distance from the neutral point 5 mm, joint height 0.1 mm, eutectic Sn–Pb, mean joint temperature 5 °C, 15.1 cycles per day (about 5,515 a year). 11 ppm/K corresponds to germanium-based cells on FR-4.

The combination of low heat capacity and the germanium-on-FR-4 mismatch puts rigidly soldered large parts in a regime where joint life is comparable with mission duration. Mitigations include CTE-tailored substrates, compliant silicone attachment with flexible interconnects, smaller distances from the neutral point and thermal-cycling qualification with margin[16].

Solder and finish metallurgy

High-reliability space soldering is built around eutectic and near-eutectic Sn–Pb[17], whereas PocketQubes rely on commercial parts with lead-free finishes. Three failure modes deserve attention. Tin whiskers grow spontaneously from tin-rich finishes under compressive stress; in vacuum a whisker short can sustain a metal-vapour arc[18]. Tin pest, the transformation of β-tin into brittle α-tin below 13.2 °C, is slow and inhibited by lead, bismuth and antimony, but cold-biased interiors and lead-free joints make it worth noting. Gold dissolving from thick finishes forms brittle AuSn₄, which is why space soldering standards require de-golding.

Outgassing screening

ECSS-Q-ST-70-02C heats a sample to 125 °C for 24 h in vacuum with a collector at 25 °C and accepts it when the recovered mass loss (RML, total mass loss minus water vapour regained) is below 1.0% and the collected volatile condensable material (CVCM) below 0.10%[19]. ASTM E595 applies the same thresholds to total mass loss[20]. PQ-Gen-04 states TML and CVCM limits but cites the ECSS document, so it is the stricter reading and can reject hygroscopic materials that ECSS accepts[1].

Material	TML	CVCM	WVR	RML	Screening	Erosion yield (10⁻²⁴ cm³/atom)	Recession, tumbling face (µm/yr)
FR-4 laminate	0.22	0.00	0.09	0.13	Passes ECSS and ASTM	—	—
G-10 laminate	0.52	0.00	0.16	0.36	Passes ECSS and ASTM	—	—
Polyimide–glass laminate	0.71	0.00	0.40	0.31	Passes ECSS and ASTM	—	—
PEEK, moulded	0.14	0.00	0.05	0.09	Passes ECSS and ASTM	2.99	16
PEI (Ultem), moulded	0.40	0.00	0.32	0.08	Passes ECSS and ASTM	> 3.31	18
ULTEM 1010, FDM (XY)	0.55	0.02	0.39	0.16	Passes ECSS and ASTM	—	—
Polypropylene filament	0.24	0.05	0.03	0.21	Passes ECSS and ASTM	—	—
Polyimide film (Kapton)Erosion-yield reference	0.97	0.00	0.80	0.17	Passes ECSS and ASTM	3	16
PET film (Mylar)	0.24	0.01	0.18	0.06	Passes ECSS and ASTM	3.01	16
PTFE	0.01	0.00	0.02	−0.01	Passes ECSS and ASTM	0.142	0.8
ABS	—	—	—	—	Grade-dependent	1.09	5.8
Polyamide 6	—	—	—	—	Grade-dependent	3.51	19

TML, CVCM, WVR and RML = TML − WVR in per cent; limits 1.0% mass loss and 0.10% CVCM (ECSS-Q-ST-70-02C on RML, ASTM E595 on TML). Recession = erosion yield × fluence, using the ISS ram fluence of MISSE-2 (2.1 × 10²¹ atoms/cm² a year) divided by four for a tumbling face; at 500–550 km the fluence is lower.

Data come from the NASA outgassing database[21] and, for erosion yields, from the MISSE-2 PEACE experiment on the International Space Station[22].

Printed polymers

FDM parts are anisotropic: ULTEM 9085 loses about 40% of its tensile strength between the in-plane and build directions, so the build orientation of a part must be recorded with its outgassing data. Porous printed parts absorb moisture, which inflates mass loss and makes bake-out effective. ABS and PLA are fine for engineering models but must be replaced or qualified individually for flight[1].

Atomic oxygen

External polymer surfaces erode under atomic oxygen; the recession depth is the erosion yield times the fluence. In ram on the ISS, PEEK recedes by about 60 µm a year; a randomly tumbling face sees about a quarter of the ram fluence, and the fluence at 500–550 km is lower still and depends on solar activity[1].

Design and verification at a glance

The magnetic side of materials selection, from AlNiCo magnets to amorphous dampers, is covered in the PMAC entry.