How it is calculated
Database search. The page loads NASA’s table once (13,582 rows, 1.4 MB, about 360 kB as sent compressed) after the rest of the page has loaded, and filters it in the browser. A test is within the limits when
both as published (to 0.01 %). On the ESA basis, the mass loss is \(\text{RML} = \text{TML} - \text{WVR}\), and both comparisons are strict (<), as ECSS-Q-ST-70-02C writes them. A test missing TML or CVCM is shown as incomplete and never counted as passing.
Gas load. Each surface \(i\) of area \(A_i\) outgasses at a rate that falls as a power of the pumping time \(t\), the empirical law found for nearly all unbaked materials:
For a rate \(q_\text{ref}\) quoted at another time \(t_\text{ref}\), the rate after 1 h on the same law is \(q_1 = q_\text{ref}\,(t_\text{ref}/1\ \text{h})^{n}\); several presets below were converted this way. A constant hydrogen rate from baked metal is \(n = 0\).
Pressure. Hours into a pump-down, the air in the volume is long gone (it is pumped with the time constant \(V/S_\text{eff}\), a fraction of a second for a small chamber on a turbo pump), so the pressure follows the gas load, as in the pump-down equation with \(dp/dt\) negligible:
Every term of \(Q(t)\) falls or stays constant, so the time to reach a target pressure is the single root of \(Q(t) = S_\text{eff}\,(p_\text{target} - p_\text{pump})\), found by bisection in \(\ln t\). The speed that holds the target at time \(t\) is \(Q(t)/(p_\text{target} - p_\text{pump})\). Units: 1 mbar·L/(s·cm²) = 1000 Pa·m/s, and 1 Torr·L/(s·cm²) = 1333.2 Pa·m/s.
Typical values
The presets, with their sources. For unbaked materials the scatter between nominally identical samples is an order of magnitude or more, and the cleaning, the humidity the parts were stored in, and how long they stood open matter as much as the material. Rates in mbar·L/(s·cm²) at room temperature.
| Material and condition | Rate after 1 h | n | Source and basis |
|---|---|---|---|
| Metals (stainless steel, aluminum, copper, titanium), cleaned, unbaked | 3 × 10⁻⁹ | 1 | CERN design rule q = 3 × 10⁻⁹/t[h], mostly water (Chiggiato, eq. 13 and Table 3; Baglin, eq. 11) |
| Stainless steel, literature average | 1.9 × 10⁻⁷ | 1 | Averages of published values at 1 h (Edwards application note, from Grinham and Chew); n ≈ 1 for metals and glass |
| Aluminum, literature average | 3.0 × 10⁻⁷ | 1 | |
| Copper, literature average | 2.3 × 10⁻⁸ | 1 | |
| Mild steel, fresh | 6.2 × 10⁻⁷ | 1 | |
| Titanium / borosilicate glass (Pyrex) | 1.0 × 10⁻⁸ / 9.9 × 10⁻⁹ | 1 | |
| Stainless steel baked 24 h at 150 °C (H₂) | 3 × 10⁻¹² | 0 | Chiggiato, Table 3 (Cu measured at CERN; steel and Al from his refs. 13–16) |
| Aluminum alloys baked at 100–200 °C, about 20 h (H₂) | 1 × 10⁻¹³ | 0 | |
| OFS copper baked 24 h at 200 °C (H₂) | 3 × 10⁻¹⁴ | 0 | |
| 316LN vacuum-fired at 950 °C, then baked at 300 °C (H₂) | 6 × 10⁻¹⁵ | 0 | Baglin |
| Fluoroelastomer (Viton, FKM), unbaked | 1.1 × 10⁻⁶ | 0.7 | Edwards (1 h); n chosen to give 2.2 × 10⁻⁷ at 10 h, against Baglin’s 2 × 10⁻⁷ |
| Fluoroelastomer, after bake-out | 5 × 10⁻¹⁰ | 0 | Baglin |
| Neoprene | 4.0 × 10⁻⁵ | 0.6 | Edwards (1 h); n from Chiggiato’s 10⁻⁵ at 10 h |
| PTFE / PVC (stored at 95 % humidity) | 1.4 × 10⁻⁶ / 3.2 × 10⁻⁶ | 0.5 | Edwards (1 h); n = 0.5, diffusion from the bulk (polymers 0.4–0.8, Grinham and Chew) |
| PEEK, 2 mm thick, unbaked | 4.7 × 10⁻⁷ | 0.5 | Chiggiato: 500 times the metal rule (1.5 × 10⁻⁷) at 10 h, slope −0.5 after 3 h |
| Kapton HN film, 0.125 mm, unbaked | 6.4 × 10⁻⁸ | 0.5 | CERN, Sammartano et al. (1.3 × 10⁻⁸ and 1.6 × 10⁻⁶ at 24 h), back-extrapolated as 1/√t |
| Vespel SP-1, 1.3 mm, unbaked | 7.8 × 10⁻⁶ | 0.5 |
The cleaned-metal rule and the literature averages differ by a factor of 60 for stainless steel. The averages include as-received and poorly cleaned surfaces, so use the CERN rule for parts cleaned for vacuum and the averages as a pessimistic case. Thin films such as Kapton deplete their water within hours: after the bend, the rate falls much faster than 1/√t (Baglin gives 5 × 10⁻¹⁰ after 100 h for 0.1 mm), so the preset overestimates it at long times. ISO/TS 20177:2018 sets out how outgassing rates should be measured and reported; quote the pumping time with any rate you measure.
Worked example
The defaults: a chamber with 5400 cm² of cleaned, unbaked stainless steel walls (a 30 cm cube), 20 cm² of Viton O-ring surface facing the vacuum, and 50 cm² of PEEK insulators, pumped at an effective 200 L/s.
- After 10 h the steel gives 3 × 10⁻⁹/10 × 5400 = 1.62 × 10⁻⁶ mbar·L/s, the Viton 1.1 × 10⁻⁶ × 10−0.7 × 20 = 4.39 × 10⁻⁶, and the PEEK 4.7 × 10⁻⁷ × 10−0.5 × 50 = 7.43 × 10⁻⁶ mbar·L/s.
- The total, Q = 1.34 × 10⁻⁵ mbar·L/s, holds the chamber at 1.34 × 10⁻⁵/200 = 6.7 × 10⁻⁸ mbar. The 70 cm² of polymer, 1.3 % of the area, supplies 88 % of the gas.
- To reach 10⁻⁸ mbar takes 253 h, about 10.5 days. The steel alone would get there in 8.1 h: 3 × 10⁻⁹ × 5400/(200 × 10⁻⁸) = 8.1. Holding 10⁻⁸ mbar at 10 h would take 1340 L/s, a much larger pump; replacing the PEEK with metal parts and baking the seals is cheaper.
- In Part 1, a search for PEEK finds 25 tests, 24 within the NASA limits. PEEK is a sound low-contamination choice, and still the largest gas source here. The two criteria measure different things.
Assumptions and limits
- Power-law rates. \(q \propto t^{-n}\) is an empirical fit that holds over a few decades of time. Water from metals follows 1/t from minutes to weeks; diffusion from polymers follows 1/√t until the part is depleted, after which it falls faster (thin films within hours). The calculator does not model that bend, so at long times it overestimates thin polymer parts.
- Room temperature. Outgassing rises steeply with temperature; a heated stage or a lamp near a polymer part changes everything. Bake-out lowers the later rates by orders of magnitude, and the baked presets assume it has been done.
- Each surface independent. Readsorption on the other walls, permeation of air and water through elastomer seals (about 10⁻⁷ mbar·L/s of water through a 6 cm Viton O-ring with a 5 mm cord, once it reaches steady state after about two months), leaks, and virtual leaks (trapped volumes behind screws and welds) are not included.
- One pumping speed. Pumps have different speeds for different gases, and a turbo pump’s speed and compression are lowest for hydrogen. Use the speed for the gas that dominates, usually water.
- E595 numbers are not rates. TML and CVCM come from a 125 °C test and cannot be converted into a room-temperature rate. NASA’s guide notes that formulations change and recommends testing each lot for critical uses. Manufacturer codes are shown as NASA lists them; the current NASA site does not publish the code list.
References
Primary sources and standards
- NASA Goddard Space Flight Center, Engineering and Technology Directorate, “Outgassing Database” and “Outgassing Database User Guide,” etd.gsfc.nasa.gov (retrieved 6 November 2026) (all 13,582 test results shown here, the test procedure, the 1.0 % TML and 0.10 % CVCM criteria, cure notation, and the caution on batch testing). The data are a US Government work; NASA does not endorse this page.
- W. A. Campbell Jr. and J. J. Scialdone, “Outgassing Data for Selecting Spacecraft Materials,” NASA Reference Publication 1124, Revision 3 (November 1990) (the printed compilation the database continues).
- ASTM E595-15(2021), Standard Test Method for Total Mass Loss and Collected Volatile Condensable Materials from Outgassing in a Vacuum Environment, ASTM International (125 °C, below 7 × 10⁻³ Pa, 24 h, 25 °C collector; a screening method, not valid for computing contamination of a system).
- ECSS-Q-ST-70-02C, Space product assurance — Thermal vacuum outgassing test for the screening of space materials, ESA Requirements and Standards Division (15 November 2008), ecss.nl (RML = TML − WVR; limits RML < 1.0 % and CVCM < 0.10 %).
- P. Chiggiato, “Outgassing properties of vacuum materials for particle accelerators,” lecture notes, CERN Accelerator School: Vacuum for Particle Accelerators (Glumslöv, 2017), arXiv:2006.07124 (2020) (the 3 × 10⁻⁹/t rule for cleaned metals, eq. 13; rates of baked metals and neoprene, Table 3; PEEK’s rate and 1/√t decay, Fig. 30; Viton permeation).
- V. Baglin, “Vacuum in accelerators for mechanical & materials engineering,” CERN Accelerator School: Mechanical & Materials Engineering for Particle Accelerators and Detectors (Sint-Michielsgestel, 2024), arXiv:2601.11361 (the same metal rule, eq. 11; Viton unbaked and baked; Kapton after 100 h; vacuum-fired 316LN).
- S. G. Sammartano, I. Wevers, G. Bregliozzi, and P. Chiggiato, “Outgassing rates of PEEK, Kapton and Vespel foils,” CERN ATS note, draft of 13 October 2021, attached to I. Wevers, “Outgassing of Polymers,” CERN workshop Materials and Engineering Technologies for Particle Accelerator Beam Diagnostic Instruments (21–23 June 2021), indico.cern.ch (Table 2, rates after 24 h by thickness; 1/√t decay and its later bend).
- R. Grinham and A. Chew, “A review of outgassing and methods for its reduction,” Appl. Sci. Converg. Technol. 26, 95–109 (2017), doi:10.5757/ASCT.2017.26.5.95 (averaged literature rates, Table 4, and the time exponents of metals, polymers, and porous surfaces); converted to mbar at 1 h in Edwards, “Outgassing,” application note 3601 2171 01 (2021).
- R. J. Elsey, “Outgassing of vacuum materials — I,” Vacuum 25, 299–306 (1975), doi:10.1016/0042-207X(75)90730-7, and “— II,” 25, 347–361 (1975), doi:10.1016/0042-207X(75)91653-X (the classic review of outgassing measurements, mechanisms, and their time dependence).
- P. A. Redhead, “Modeling the pump-down of a reversibly adsorbed phase. I. Monolayer and submonolayer initial coverage,” J. Vac. Sci. Technol. A 13, 467–475 (1995), doi:10.1116/1.579381 (why water desorption from a surface with a spread of binding energies falls about as 1/t).
- P. A. Redhead, “Recommended practices for measuring and reporting outgassing data,” J. Vac. Sci. Technol. A 20, 1667–1675 (2002), doi:10.1116/1.1496783 (why a rate needs its pumping time, temperature, and history).
- ISO/TS 20177:2018, Vacuum technology — Vacuum gauges — Procedures to measure and report outgassing rates.
- J. F. O’Hanlon, A User’s Guide to Vacuum Technology, 3rd ed., Wiley (2003), doi:10.1002/0471467162, ch. 4 (gas release from solids: desorption, diffusion, and permeation, and their time dependence).
Free introductions
- P. Chiggiato, “Outgassing properties of vacuum materials for particle accelerators,” arXiv:2006.07124 (the most complete free treatment: water, hydrogen, polymers, bake-out, and how rates are measured).
- Edwards Vacuum, “Outgassing,” application note 3601 2171 01 (2021), edwardsvacuum.com (two pages: the gas sources, the power law, and a table of rates at 1 h).
- NASA Goddard, “Outgassing Database User Guide,” etd.gsfc.nasa.gov (how the E595 test works and how to read the table).