NASA GRC Glenn Extreme Environments Rig

NASA. Public domain (NASA / US government work).

The Glenn Extreme Environments Rig is a 12 t, 800 L stainless steel pressure vessel that holds Venus surface temperature, pressure and gas chemistry indefinitely, with a nine stream mixing system that injects trace species at parts per billion [1]. It is the facility where NASA’s high temperature silicon carbide electronics, Venus sensors and candidate lander materials are exposed to the atmosphere they will have to survive, rather than to temperature alone [5]. GEER is the largest of several extreme-environment rigs at NASA Glenn, which also runs cryogenic chambers built to hold the hydrogen, hydrocarbon and nitrogen atmospheres of the outer planets and Titan; GEER is the one built for a hot, corrosive, high pressure atmosphere rather than a cold one [7].

Laboratory data

Parameter	Value
Operator	NASA GRC; supported by the Planetary Science Division [1]
Location	Cleveland, Ohio, United States
Commissioned	November 2014; research operations from spring 2015
Type	High pressure, high temperature reactive-atmosphere chamber
Floor area	Not published
Capabilities	GEER vessel, gas system, Mini GEER
Simulant or terrain	Not applicable; the chamber holds a gas mixture, not a bed
Instrumentation	In-line four-column gas chromatograph; on-site mass spectrometer
Ground truth	Gas chromatography and mass spectrometry during the run, independent of the charge
Fidelity limits	No solar flux, no wind loading, no dynamic descent profile
Access	Two-phase process; 4 to 8 weeks planning, 30 workdays minimum per test
Cited by	SiC JFET-R integrated circuits and packaging for Venus landers [3][4][6]; alloy and coating corrosion screening [8]; Venus lander concept studies [10]

Access is a two-phase process. Phase I test planning requires a GEER Test Requirements Document, and the operator quotes 4 to 8 weeks to settle scope of work, cost estimate and test agreement; Phase II is execution [1]. Proposed articles go to a Material Review Board of scientists and GEER operations staff, which may reject materials liable to react hazardously with the vessel atmosphere. The minimum test process is 30 workdays, about six weeks, made up of 3 days bolt-up, 10 days leak check and purge, 4 days fill and heat-up, time at conditions per requirements, 10 days cool-down, plus sample loading and unloading [1]. All GEER test photographs are non-restricted and processed for public release unless the customer flags proprietary content, and no fee schedule is published. Over 20 tests had been conducted in the chamber by 2024 [1].

Capabilities

GEER vessel

NASA/GRC/Bridget Caswell. Public domain (NASA / US government work).

Parameter	Value
Working volume	36 in diameter x 48 in deep (0.91 x 1.22 m), about 800 L [1]
Test article limits	Placement envelope 24 in diameter x 36 in; no article mass limit published
Vacuum	Rated upward to 1500 psig, about 103 bar
Temperature	Ambient to 1000 F (about 530 C); ramp 7 C/h [4]
Illumination	Provision for two 4 in optical ports at opposing ends [5]
Slope	Not applicable
Gravity offload	Not applicable
Instrumentation	Calibrated pressure and temperature data system; 48 feedthrough wires maximum [1]

GEER is an ASME stamped pressure vessel of 304 stainless steel, 12 t, originally designed with the engineering limits of 1000 F and 1500 psi needed to reach Venus surface conditions with margin [1]. The bolted head withdraws on a carriage to open the full bore, so large articles go in through the main door while small hardware and materials go through the smaller ports [5]. The wall is 2 in thick, the bolted end 9.5 in and the welded end 6.75 in, and the interior is lined throughout with Inconel 625 to limit attack by the harsh gas mixtures; gas plumbing is 304 stainless except in the high temperature regions adjacent to the vessel, where it changes to Inconel 625 with Monel fittings. Nine flange ports cross the fixed and bolted heads, two of 4 in and seven of 3 in; four of the 3 in ports carry vessel hardware and instrumentation, leaving three 3 in and two 4 in ports free for customer feedthroughs [1].

Venus surface tests run at 460 +/- 5 C and 92 bar, held through a run by a re-boost pumping system, and uniformity across the 800 L volume is not published [4]. Thermal operation is slow by design: at 7 C per hour an article spends roughly 60 h transiting from ambient to Venus conditions, and a Venus-conditions test spends four days filling and heating and ten days cooling around whatever dwell the customer requires.

Fixturing determines what actually fits. A shelf gives a 36 x 25 in area at about 12 in height for many small articles; half-pipe supports of 3 in and 4 in diameter run to 42 in long; arch supports take large articles bottom-supported within the article envelope; and electrical feedthrough platforms carry small samples on alumina substrate [1]. Standard alumina crucibles run 200 x 100 x 25 mm down to 50 x 20 x 20 mm, with circular 45 x 45 mm and 25 x 30 mm [1]. Custom fixturing can be fabricated, and multiple fixtures may not fit in one test. Electrical feedthroughs are proven at plus or minus 24 VDC at low single-digit mA per wire, with a higher- capacity feedthrough demonstrating a stable 15 VDC at 1 A for three days at Venus surface conditions [4].

Gas system

Parameter	Value
Working volume	Nine independent streams feeding the 800 L vessel [5]
Simulant or terrain	CO2, N2 and up to eight trace gases at parts per billion, ppm accuracy [1]
Instrumentation	Thermal mass flow controller per stream; in-line four-column gas chromatograph [5]

The gas system is the part that distinguishes GEER from a furnace. A bank of cylinders feeds a mixing system delivering up to nine independent streams at controlled concentration, each metered by its own thermal mass flow controller from a bottle that may itself hold a premixed set of gases, so effectively any planetary atmosphere can be made up within the limits of the containing materials, and the composition can be adjusted while the test is running [5]. A boost pump recharges the chamber at pressure, and a scrubber, gas containment, programmable control and gas analytics complete the installation.

The Venus mixture published for the facility’s system test is 96.5 percent CO2 and 3.4 percent N2 with SO2 at 130 ppm, COS at 27 ppm, CO at 15 ppm, H2O at 30 ppm, HCl at 0.5 ppm, NO at 5.5 ppb and HF at 5 ppb [1].

Mini GEER

Parameter	Value
Working volume	5 in internal diameter x 12 in, about 4 L (unverified, see below)
Test article limits	Materials coupons and small components (unverified, see below)
Vacuum	Vacuum to 2706 psia (unverified, see below)
Temperature	Ambient to 950 F, about 510 C (unverified, see below)

Mini GEER is the small-volume companion vessel, roughly 200 times smaller than GEER, used to screen materials and small components at Venus surface conditions without committing the 30 workday cycle a GEER run requires [1]. The 2024 users guide describes only the main vessel [1].

Why it exists

GEER exists because of a specific gap. A 2012 Venus Exploration Analysis Group survey of laboratories able to reach Venus surface temperature and pressure with at least CO2 and N2 found nine facilities [5]. The largest was just over 1 ft3 of internal volume and only four exceeded 0.1 ft3; only seven reached both Venus surface temperature and pressure, the largest of those 0.5 ft3; only the JPL and Goddard chambers could handle more than CO2 and N2; and none could handle all the known Venus atmospheric species including HF, HCl, H2S and H2SO4. The survey’s conclusion was that as of March 2012 there was no operational capability for full-scale testing of instruments or spacecraft systems for the Venus surface [5]. GEER’s design targets followed directly: part per billion composition control, supporting infrastructure sized for a chamber at least 7 ft in diameter, an intent to simulate the changing composition and pressure and temperature growth of a parachute descent, and flexibility to reach complex compositions at cryogenic temperatures. The supporting plant was deliberately sized to serve multiple chambers or much larger ones than the vessel installed, which is the design provision Mini GEER now occupies.

The gap GEER closed is why a 2020 Glenn concurrent engineering study of a low-cost Venus orbiter and surface lander used GEER exposure results, rather than a materials handbook, to choose the surface probe’s structural alloys and coatings; the study’s cost cap depended on flying hardware whose survival in the real atmosphere had actually been measured rather than extrapolated from a furnace test [10]. The same demand for high-temperature electronics that can survive the Venus surface for a useful mission life is carried in the 2022 planetary science decadal survey as one of the technology gaps standing between current silicon electronics and long-duration Venus in-situ science [11].

Instrumentation

Facility instrumentation records chamber pressure and temperature through a calibrated data system. Atmosphere composition is tracked through the run by an in-line four-column gas chromatograph, with an on-site mass spectrometer available; a summary of gas analysis results is provided after each test on request [1].

Recording data from an active test article is the customer’s responsibility; GEER provides the feedthroughs and helps with setup, and a suite of post-test microanalysis covering SEM, EDS, XPS, Auger and Raman sits alongside the chamber with technical personnel [1]. That division has shaped the electronics campaigns run there, which bring their own instrumentation outside the vessel, measure through the wire count the feedthroughs allow, and reach their conclusions from post-test microscopy rather than from in-test telemetry [3].

What it does not reproduce

Chemistry drifts during a run. The atmosphere the article sees is not the atmosphere that was charged. GEER’s own in-line gas chromatograph exists because composition has to be measured through the run rather than assumed, and a gas analysis summary is issued after each test [1]. The Pt/HTCC package campaign is the clearest published instance: insulation resistance degraded heavily even in the nominally inert 90 bar nitrogen phase, which the investigators recorded as unanticipated, and post-test analysis found platinum sulfide formed on the bond pad from the sulfur-bearing gases [4].

No solar flux, no wind loading, no descent profile. The vessel has no illumination source, no means of producing flow across an article, and no published means of running a dynamic descent profile in pressure and temperature; the published process holds the chamber at a single set of conditions for the dwell [1].

Ramp rate is a test condition, not an artifact to be ignored. At 7 C per hour, an article spends roughly 60 hours transiting from ambient to 460 C [4]. That transit is where the SiC multi-chip board failed, at 107 C during heat-up, long before Venus conditions were reached [3].

Endurance testing at temperature is not done here. The thousands of hours of 500 C operation demonstrated for individual SiC JFET-R integrated circuits were run in room-air ovens, not in GEER, because a GEER cycle costs a minimum of 30 workdays [6]. GEER supplies the atmosphere; the ovens supply the hours. A device qualified for 10,000 hours at 500 C in air has not been shown to survive 10,000 hours in supercritical CO2 with SO2, OCS, HCl and HF present [6]. Radiation qualification of the same SiC JFET-R chips is a third, separate track again: total ionizing dose and single-event effect testing to over 7 Mrad(Si) was run at room temperature in a particle beam, not in GEER’s atmosphere or in a high temperature oven, because GEER holds atmospheric chemistry and temperature, not radiation [9]. A chip’s Venus qualification is assembled from three facilities, none of which alone demonstrates the whole environment.

Bare material coupons do not predict a packaged part. A structured coupon survey exposed plain steels, nickel alloys, gold, ceramic coatings and bulk ceramics to the Venus mixture at 92 bar and 467 C for 10, 42 and, for coated stainless, 21 days: alpha-Al2O3, Si3N4, SiC and fused quartz showed no detectable weight change at either duration, gold was unaffected, and the steels and nickel alloys all formed double-layered oxide-sulfide scales whose thickness ordered as 304 < 310 < 316 < 1018 among the steels and beta-NiAl < G30 < 625 among the nickel alloys [8]. That ranking is a materials selection tool, not a lifetime prediction: it says which bare alloy corrodes fastest under GEER conditions, not how a bonded, packaged, electrically active device behaves, which is what the Pt/HTCC and multi-chip board campaigns below test instead [8].

Operational findings. An independent NASA Engineering and Safety Center review in 2016 identified twelve findings, most minor [2]. Four bear on the vessel itself [2]:

• No procedure existed for documenting the operational history of the 304 stainless steel vessel, which the review recommended be recorded for creep assessment at recertification given prolonged operation at elevated temperature [2].
• The wet fire suppression system in the test cell could thermally shock the vessel if inadvertently activated.
• Corrosion under the vessel insulation was not in the inspection or maintenance plan, although API-571 treats it as a concern for insulated 304 stainless operating intermittently between 60 and 204 C [2].
• There was no means to visually monitor the test cell from the control room [3].

Campaigns run there

SiC JFET-R single-chip ICs, 60 days, published 2018. Individual NASA Glenn IC Generation 10 silicon carbide junction field effect transistor circuits, under 200 transistors per chip, functioned for at least 60 days immersed in the chamber-simulated Venus surface environment without sheltering [3]. That result is what the multi-chip board work below sets out to scale up [3].

Pt/HTCC alumina package, 60 days, published 2018. A prototype platinum on high-temperature co-fired ceramic alumina package was run through three phases: Earth air, GEER nitrogen at 90 bar, and simulated Venus surface conditions, heated at 7 C/hour to 460 +/- 5 C for 60 Earth days [4]. Interconnect resistance changes seen in GEER nitrogen and in GEER Venus conditions were not reversible. Assembly 5’s insulation resistance entered the Venus phase at 45 kilohm, fell to a minimum of 721 ohm at 231 hours and had recovered only to 2.26 kilohm at 1459 hours, about 60 days; assembly 4 fell from 1.18 megohm to 19.8 kilohm across the same phase [4]. The degradation during the nominally inert 90 bar nitrogen phase was itself unanticipated [4]. Running an inert high-pressure phase before the reactive phase is what separated pressure and temperature effects from chemistry effects.

IC Generation 11 multi-chip lander board, 11 days, published 2023. A first attempt at a complete lander control and sensor digitization subsystem: an 11.5 x 11.5 cm ceramic circuit board with four interconnect layers carrying ten SiC JFET-R chips in nine ceramic packages, up to 706 transistors per chip, performing analogue to digital conversion of multiple sensor voltages into serial data [3]. The board ceased electrical operation at 107 C as the chamber heated up, well short of the 460 C, 9.4 MPa target. Post-test optical and SEM inspection found fatal dielectric cracks on one of the ten chips, in circuit-critical regions including a major power supply bus and the DAC subcircuit, correlated with dielectric hillock defects formed during deposition of the third SiO2 layer [3]. The suspected root cause was SiO2 deposition onto localized photoresist etch mask remnants, with the hypothesis that most of the cracking occurred during a 600 C die-attach air anneal rather than in the chamber. No package, board, wire bond or die-attach failure was found. The mitigation for Generation 12 is revised metal etch processing to eliminate photoresist residue and increased bond pad over-etch [3].