At the Institut für Raumfahrtsysteme (IRS) ofUniversität Stuttgart, inductively heated plasmagenerators have been developed to provide high en-thalpy plasma flows for basic thermal protectionmaterial testing foreseen for entry maneuvers inthe atmospheres of Earth, Mars, or Venus [1,2].Due to the absence of electrodes high enthalpy plas-ma states are achieved even when using reactive dri-ver gases, e.g., O2, CO2, and corresponding mixtures.During Earth reentry, the boundary layer in front ofthe thermal protection material is neither in thermalnor in chemical equilibrium. To better understandthe rather complex plasma wall interactions andhence the material behavior, inductively heated plas-ma wind tunnels are used to investigate the behaviorof candidate materials in pure plasma flows of the
main constituents of the air plasma, i.e., N2 andO2. This is of particular interest for understandingthe catalytic and oxidation behavior of candidatematerials [1,3,4].
Pure oxygen plasma flows have been characterizedso far by using different probe techniques. By mea-suring the total pressure and local heat flux on oxi-dized copper together with measurements using acavity calorimeter, the local mass-specific enthalpyhas been estimated for a reference plasma condi-tion to be h ¼ 27MJ=kg, while chemical and thermalequilibrium had to be assumed [5]. The flow hereinwas treated as fully dissociated, an assumption thatis based on calculations of the chemical equilibriumfor molecular oxygen (O2 → OþO). The probe posi-tion with respect to the generator exit was chosenas x ¼ 140mm from the generator exit in the centerof the plasma plume. Recently, the local mass-specificenthalpy of the same plasma condition at thesame position, but based on diode laser absorption
spectroscopy (DLAS) of the 3p 5P state at λ ¼777:19nm of the oxygen atom was measured [6].From the measured temperature and under the as-sumption of chemical and thermal equilibrium,the local mass-specific enthalpy was estimated toh ¼ 33:7� 3MJ=kg. In another conclusion a lowerlimit for the degree of dissociation of Δ > 0:92 wasdeduced [7].For the planned reentry capsule EXPERT (Eur-
opean experimental reentry testbed), IRS is incharge of the development and qualification of anew sensor that measures the dissociation degreeof the plasma within the boundary layer at thevehicle’s surface [8]. The sensor, named PHLUX (pyro-metric heat flux experiment), is based on measuringthe heat flux on two material samples with differentbut known catalytic properties. First experimentsusing a water-cooled laboratory model in the induc-tively heated oxygen plasma flow yielded full disso-ciation even for positions farther downstream thanthe aforementioned reference position. However, acomparison measurement based on the most impor-tant constituent of the pure oxygen flow, i.e., groundstate atomic oxygen, was lacking until now. More-over, knowledge of the local particle density of atomicoxygen and temperatures is needed to characterizethe plasma state in order to further understandthe sensor’s principle and to improve numerical cal-culation tools.Quantitative measurements of number densities
of the ground state population of major species inpure oxygen high enthalpy plasma flow have recentlybeen measured by some of us using new calibrationtechniques [9]. With respect to the development ofthe flight hardware for the EXPERT capsule, the pre-sent investigation can be considered as a cross cali-bration reference. Moreover, due to the nonintrusivelaser-induced fluorescence (LIF) technique, the un-perturbed free stream is characterized such thatthe results can be used for all future probe designs.
2. Experimental Setup
The plasma wind tunnel facility named PWK3 con-sists of a vacuum vessel 2m in diameter and 2:5min length (see Fig. 1) [10]. Optical access on both sidesof the vacuum chamber is provided in order to mea-sure and observe the plasma free jet. The plasmagenerator depicted in a sectional view in Fig. 2 ismounted on the front lid. On the rear end, the va-cuum chamber is connected to a vacuum system witha suction power of up to 250; 000m3=h at 10Pa. Thesize of the plasma wind tunnel allows flight experi-ments to be tested and qualified in real size up todiameters of about 150mm without disturbancesfrom wall interaction effects with the vacuum vesselor ambient pressure variations even at high gasmass flows.The plasma is generated inside a thin quartz tube
by a high frequency current applied to the inductioncoil around the tube (see Fig. 2). The coil has 5.5turns. In combination with four capacitors, the coil
is part of aMeissner-type resonant circuit that is con-nected to the power supply, which has a maximumelectric power of 375kW. Both the quartz tube andthe induction coil are water cooled. The working fluidis fed into the quartz tube from one side using tan-gential injection. The alternating current in the coilinduces an azimuthal electric field inside the tube.Then, electric discharge in the gas is initiated dueto inherently present free electrons. The producedplasma expands into the vacuum chamber. Theadjustable parameters are mass flow rate _m, ambi-ent pressure pa in the vacuum chamber, and anodevoltage UA of the power supplying triode. The mea-surements presented in this paper have all been per-formed at the minimum possible ambient pressure ofpa ¼ 40Pa. With increasing anode voltage, the elec-tric power fed to the resonant circuit also increases.Meanwhile, the tangential injection combined witha frequency optimization for oxygen allows anodevoltages from zero up to UA ¼ 6500V dependingon the gas mass flow [2]. There are three differentdischarge modes. At very low anode voltages, a dis-charge without glowing occurs. The second mode atlow anode voltages corresponds to a capacitive dis-charge. The switch from this second mode to the in-ductive third mode is discrete and occurs at anodevoltages between UA ¼ 3000V and UA ¼ 5000V de-pending on the mass flow rate [2]. Both dischargeconditions are dominated by the frequency of the re-sonant circuit, leading to a fast oscillating plasma.The LIF investigations performed at a repetition rate
of 10Hz cannot resolve the high frequency discharge.Hence, the measured fluorescence is a mean valueand so are the deduced densities.The experiments of the present work aim at quan-
tifying the most important constituent of the pureoxygen plasma: ground state species of atomic oxy-gen. The number densities are measured by usingLIF. The basic approach of this measurement techni-que is sketched in the energy level diagram in Fig. 3.Oxygen atoms in ground electronic state are excitedvia the 3p 3P2;1;0←2p 3P2 two-photon transition at225:5nm. The following fluorescence is measuredusing the 3p 3P2;1;0 → 3s 3S0
1 transition at 844:5nm.The excitation is performed with a pulsed laser sys-tem, and the fluorescence is detected by using aphotomultiplier and a fast oscilloscope as well asgated boxcar integrators.The alignment of the experimental setup for the
spectroscopic measurements is sketched in Fig. 1.The laser system is installed on top of the vacuumchamber. It consists of a pulsed excimer gas laserfilled with XeCl emitting at 308nm (Lambda PhysikCompEX 205). With a dye laser (Lambda PhysikScanmate 2E) and a β-barium borate (BBO) crystalfor frequency doubling, laser radiation at 225nmis attained with a maximum output of 1mJ perpulse. The laser beam is directed to the measure-ment location on the plasma jet axis in the centerof the vacuum chamber by three prisms. The focalpoint inside the vacuum vessel is adjusted by usinga two-lens focusing system.LIF is detected at right angles to the laser and
the flow directions (see Fig. 1). The fluorescence isdetected by a gated Hamamatsu R636-10 photomul-tiplier tube (PMT). The advantage of the gating is amuch lower background perturbation. The gate trig-ger itself is delivered by a trigger generator (StanfordResearch Systems DG535). On the detection side, inorder to avoid imaging errors, two achromatic lensesare used to image the fluorescence plane onto the de-tector. In front of the entrance slit, an interferencefilter is installed (L.O.T-Oriel 850FS20-50) with acentral transmission frequency of 850nm and a fullwidth at half-maximum of 20nm.
The data acquisition consists of two analog boxcarintegrators (Stanford Research Systems SR250) anda computer interface (Stanford Research SystemsSR245). Laser energy measurement is performedwith a pyroelectric energy monitor (Polytec RjP735).The beam splitter in use, splitting a small amount oflaser energy to the energy monitor, was character-ized prior to the experiments. The amount of energythat is split off is 3% of the incoming energy in thewavelength region of interest [10]. The measure-ments are observed in parallel by a fast 1GHz oscil-loscope (Gagescope 82G) in order to evaluate theeffective lifetimes of the upper state. This oscillo-scope is integrated into the data acquisition compu-ter providing the advantage of fast data acquisitionusing the PCI (peripheral component interconnect)bus of the computer system. The data acquisitionand the laser system are controlled by a LabVIEWprogram such that for each laser pulse the time in-tegrated fluorescence, the background signal, andthe time-resolved fluorescence in situ are plottedduring the experiment and finally stored for eachlaser pulse separately.
3. Theory
The use of rate equations to describe signal intensi-ties obtained by unsaturated two-photon excitationprocesses is well described in the literature [11–13].In a simple model neglecting both depletion of theground state and photoionizing effects, the measuredfluorescence signal S as a function of the numberdensity in the ground state n0 can be written as
S ¼ Ω4πVcη
ℏω21
ℏω202
A21
AþQE2
L
A2L
σð2Þω n0
ZF2ðtÞdt; ð1Þ
where Ω stands for the observed solid angle, Vc forthe observed fluorescence collecting volume, η forthe transmission efficiency of the detection system,A21 for the Einstein coefficient of the consideredfluorescence channel, A ¼ P
k A2k is the transitionprobability of the upper level with respect to allpossible fluorescence channels, Q is the quenchrate of the excited level, EL is the laser energy, ALis the laser beam area, and FðtÞ is the temporallaser profile. The temporal profile FðtÞ of the laserpulse is measured by using a fast photodiode. Fig-ure 4 shows two laser pulses as examples along witha Gaussian fit. The analysis of more than 100 pulsesshows that the temporal profile of the laser pulse canbe assumed constant [10]. The measured laser pulseduration is τp ¼ 13:85� 1ns. The integral of thesquare of the profile is then calculated from a Gaus-sian fit of the observed profile to
RF2ðtÞdt ¼ ð ffiffiffiπp τpÞ−1.
The dependency of the LIF signal on the laser inten-sity is measured. The energy is therefore varied byturning the frequency doubling crystal and by vary-ing the discharge voltage of the pumping excimerlaser while holding the wavelength at the maximumabsorption frequency. The chosen laser intensity waslow enough to ensure that photolysis, stimulatedFig. 3. (Color online) Energy level diagram for atomic oxygen.
emission, and photoionization were negligible so thatthe LIF signal depended quadratically on laser in-tensity (see Fig. 5).The fluorescence quantum yield A21=ðAþQÞ that
accounts for fluorescence signal losses due to non-radiating deexcitation is estimated in the presentcase by time resolved measurement of the fluores-cence. The effective lifetime τLT is derived from theexponential decay of the fluorescence signal andcorresponds to the fluorescence quantum yield inthe form
A21
AþQ¼ A21 τLT: ð2Þ
The coefficient of spontaneous emission for atomicoxygen is taken from the literature to be A21 ¼ 2:89 ×108 s−1 [14].The LIF setup has been calibrated using com-
parative two-photon absorption LIF measurementson xenon gas [9,15,16]. A low-pressure gas cell was
therefore installed at the position of the plasmawind tunnel experiments. The calibration is basedon a reference measurement with xenon. A prerequi-site for the application is that both excitation areperformed with the same spatial, spectral and tem-poral intensity distribution of the laser radiation.Then, the need to know the intensity distributionexplicity vanishes. When the two resonance frequen-cies (xenon and oxygen) are close, the condition isbest fulfilled. The different fluorescence emissionwavelength can be accounted for by using differentinterference filters. Thus, a calibration using aknown xenon density can be used to quantify theatomic oxygen number density. Equation (1) holdsfor both the measurements on atomic oxygen andxenon. Relating this equation together with the cor-relation for quenching and laser pulse duration forboth measurements lead to
n0
����O
¼ Sη;OSη;Xe
E2L;Xe
E2L;O
ηXeηO
A21 τLT
����Xe
1A21τLT
����O
σð2Þω;Xe
σð2Þω;O
gðΔωXeÞgðΔωOÞ
n0
����Xe;
ð3Þwhereas geometric and detection parameters havebeen kept unchanged. The normalized line is denotedgðΔωÞ. The ratio of the absorption cross sectionshas been measured by Niemi et al. [11] to beσð2Þω;Xe=σ
ð2Þω;O ¼ 0:36� 0:18.
The translational temperature is evaluated fromthe full width at half-maximum of the measured ab-sorption line. Under the low-pressure conditionsfound in plasma wind tunnels for reentry research,pressure and Stark broadening can be neglected[17]. In the case of two-photon transitions, the linebroadening due to temperature Doppler effectsΔλtemp depends on the instrumental linewidthΔλinstr and the total linewidth which is measuredas Δλtot in the form
The experiments were conducted in an operationalcondition of the facility that allows the input powerto be varied while mass flow and ambient pressure inthe vessel are held constant. Data were acquired forsix different power conditions defined by adjustingthe anode voltage. The measurement conditions
Fig. 4. Temporal profile of the laser pulse.
Fig. 5. Quadratic dependence of the fluorescence on the laserenergy (logarithmic scale).
are summed in Table 1. All plasma states are steady-state and easily reproducible.
A. Dissociation Degree
In Fig. 6, an exemplary result of one excitation scanover the atomic absorption line is shown. Each pointcorresponds to the average of ten laser pulses and theerror bar indicates the standard deviation. Addition-ally, a Gaussian fit is plotted from which the densityand temperature values are derived. Due to the fre-quency behavior of the generator, two points of theprofile could not be analyzed (indicated by zero fluor-escence in Fig. 6). Nevertheless, an absorption profilecan be clearly identified, and the fine structure com-ponent (at λ ¼ 225:542nm) has been accounted for byapplying a second Gaussian.The absolute number density of ground state
atomic oxygen is obtained at a distance of x ¼140mm from the generator exit plane. The plasmaflow is treated as a turbulent free jet [18]. This as-sumption is strengthened by radial total pressuremeasurements [10]. A strong interaction of the plas-ma flow with the ambient air would result in measur-able effects in total pressure at the edges of theplasma free stream, which is not observed [5]. Hence,it is assumed that the static pressure in the flowcorresponds to the ambient pressure, and the total
number density can be derived from this pressureinformation. The dissociation degree of oxygen isdefined as
Δ ¼ nO
n¼ nO
pa=kT; ð6Þ
where nO is derived from the LIFmeasurements [19].The overall number density is calculated by usingthe static pressure and the measured translationaltemperature of atomic oxygen. Thus, thermal, butnot chemical, equilibrium, is assumed. The dissocia-tion increases with higher anode voltage, i.e., electricpower (Fig. 7). However, up to 5000V the dissociationdegree is below 0.5. The following sudden increasecorresponds to the characteristics of the facility.Up to 5000V, the capacitive discharge mode is estab-lished. The low power at the anode leads to low ex-citation and a low power plasma and hence to a lowdissociation of the oxygen molecules. After switchingto the inductive discharge mode, the coupling of thegenerator power into the plasma is more efficient,and for high power, condition 6 (Pel ¼ 110kW), fulldissociation is reached. An astonishing electromag-netic effect can be seen from the total pressure mea-surements. In the inductive discharge mode, thetotal pressure is about three times higher than inthe capacitive discharge mode. The Lorentz forces
Fig. 6. LIF signal as a function of laser excitation wavelength( _m ¼ 3 g=s, pA ¼ 26Pa, UA ¼ 6300V).
Fig. 7. Local dissociation degree derived from LIF measure-ments versus anode voltage of the facility as well as measuredtotal pressure.
Table 1. Test Conditions and Deduced Mach Number Ma
Condition No. Anode Voltage UA (V) Power Pel (kW) Total Pressure ptot(hPa)ptotpa
generated by the high frequency electromagneticfield obviously impresses a radial force on the plasmaflow toward the center within the tube (see Fig. 2).This leads to a nozzle effect which explains the sig-nificant increase in total pressure. As can be seenfrom Table 1, supersonic conditions arise after theswitch, which strengthens the explanation of a noz-zle effect due to the inductive heating of the gas.In the past, the dissociation degree for condition 6
has been determined by various methods. The disso-ciation degree obtained from the LIF measurementsis Δ ¼ 1:04, which is interpreted as full dissociation,taking into account the accuracy of the LIF measure-ments. In the enthalpy calculation section, the dis-sociation is set to 1 in order to fulfill the physicalmeaning of dissociation. Figure 8 summarizes theresults from the different measurements applied toestimate the dissociation degree of condition 6.The value deduced from heat flux and calorimeter
measurements is estimated under the assumption ofchemical and thermal equilibrium as well as a strongshock approximation [5]. The measurements havebeen analyzed for two different axial distances x fromthe generator exit, and at a distance of x ¼ 180mmfrom the generator exit full dissociation has been es-timated. Recent experiments that were performedfor the development of the flight experiment pyro-metric temperature measurements on different ma-terials with different, but known, catalytic properties[8] together with the catalytic recombination theorydeveloped by Goulard, have also been used to esti-mate the dissociation degree of the pure oxygen plas-ma flow [20]. This setup has been used at differentdistances from the generator exit, while full disso-ciation (Δ ¼ 1) has been reached at x ¼ 210mm forthe reference plasma condition mentioned above.Finally, a lower limit of the dissociation degree wasestimated in a previous investigation from transla-tional temperature measurements using DLAS [7].The authors concluded from these measurements,
under the assumption of chemical and thermal equi-librium at x ¼ 140mm, that Δ ≥ 0:92. The resultswithin this paper complete the tendency observedfrom the cavity calorimeter as well as the PHLUXsensor results. Because the LIF measurements con-sider chemical nonequilibrium, these results seemmore realistic. Moreover, the LIF measurementsneed fewer general assumptions. The dissociationdegree is deduced from free stream; i.e., the strongshock approximation is not needed, which can resultin an error of 30%. The boundary layer theory fromGoulard [20] strongly depends on the catalycity ofthe different materials that are known with an accu-racy of 30%. Taking into account these errors due tothe needed assumptions, the values from LIF mea-surements, with a LIF error of 35%, neverthelesslead to the most reliable result. Although the plasmaflow reaches full dissociation and therefore the influ-ence of chemical nonequilibrium is lowered, the re-sults from the LIF measurements can be used as across-calibration source for the development of theflight experiment.
B. Local Mass-Specific Enthalpy
The local mass-specific enthalpy is derived using thetemperature and density values deduced from thespectroscopic measurement. The enthalpy can eitherbe calculated using an appropriate probe techniqueor heat balancing methods, but here it is calculatedby summing up the internal energies and the energydue to the high flow velocity [21,22]. This methodthus also offers the additional possibility of in-vestigating the distribution of the enthalpy on thedifferent degrees of freedom. The translational tem-perature is assumed to be the equilibrium tempera-ture, i.e., thermal equilibrium is assumed, whichseems to be reasonable taking into account that few-er than ten collisions are needed for rotational ortranslational thermal equilibrium under these flowconditions [19]. Hence, the rotational degree of free-dom is treated as fully excited. For the pure oxygenplasma flow, the enthalpy can be then expressed as
h ¼ 12v2 þ ξO2
cprot;transT þ ξOcprot;trans
T þ ξO2hvib
þ ξOðhO þ hel;OÞ; ð7Þ
where the indices O and O2 stand for the atomic andmolecular oxygen in the flow, respectively. Plasmavelocity is estimated from total pressure measure-ments [23]: for a supersonic plasma flow the Machnumber can be derived from the well-knownRayleigh–Pitot equation in the form
ptot
pa¼
�γ þ 12
Ma2�γ=ðγ�1Þ� 2γ
γ þ 1Ma2 −
γ − 1γ þ 1
�1=ð1−γÞ
:
ð8Þ
The first two conditions, however, were subsonic.
Fig. 8. Local dissociation degree derived from LIFmeasurementscompared with PHLUX and probe measurements ( _m ¼ 3 g=s,pA ¼ 26Pa, UA ¼ 6300V).
The isentropic exponent γ is taken from calculationsof Andriatis and Sokolowa [24](see Fig. 9). Bottin [25]also calculated the isentropic exponent for induc-tively heated plasma wind tunnel flows leading toa similar result . Note that a variation of the isentro-pic exponent leads to a variation in the estimatedMach number of less than 15%. Knowing the isentro-pic exponent and by using the measured tempera-ture, the flow velocity under the assumption ofthermal equilibrium can be calculated from
v ¼ Ma⋅a ¼ MaffiffiffiffiffiffiffiffiffiγRT
p; ð10Þ
where a stands for the speed of sound and Ma forthe local Mach number. Figure 10 shows the mea-sured translational temperature and the estimated
velocity. As expected, the same qualitative resultis seen as discussed for the total pressure and disso-ciation degree measurements. However, the increasein translational temperature seems to increasemore homogeneously with increasing anode volt-age. Furthermore, a maximum temperature of T ¼3500K is already reached at UA ¼ 6000V. Thisvalue is not considerably increased with the increaseto UA ¼ 6300. The maximum velocity reached is al-most 4000 m=s. Moreover, from the measured tem-perature it can be seen that the assumption of fullrotational excitation is justified.
However, since the amount of oxygen moleculeshas not been measured yet and the vibrational modeof the molecules has to be accounted for, an assump-tion for the degree of vibrational excitation of the mo-lecules has to be made. The enthalpy due to vibrationexcitation is
hvib;O2¼
ℜMO2
θvib;O2
eθvib;O2=T
− 1: ð11Þ
In this formulation the molecule is treated as a har-monic oscillator [19]. As will be seen from the experi-ments, even an assumption of vibrational fullexcitation would yield a vibrational partition of thefree stream enthalpy of only < 5%; i.e., in the worstcase when the oxygenmolecules are not vibrationallyexcited, the measured free stream enthalpy is low-ered by a maximum of 5%.
Figure 11 shows local mass-specific enthalpy as afunction of the anode voltage as well as the distribu-tion of the enthalpy on the different modes, i.e., ther-mal, kinetic, and chemical. It is seen that most of theenthalpy is stored in the chemical mode. This is notastonishing since the concept of plasma wind tunnelsis to generate high enthalpy by high temperature. Inreal flight, of course, the high enthalpy is due to thehigh velocity of the vehicle. As already discussedabove in Section 2, the generator’s behavior gives
Fig. 9. Calculated isentropic exponents for different ambientpressures according to Andriatis and Sokolowa [24] (1atm ¼1013:25hPa).
Fig. 10. Measured translational temperature and derived freestream velocity.
Fig. 11. Local mass-specific enthalpy and enthalpy distributionin thermal, kinetic, and chemical modes as a function of the anodevoltage of the plasma generator.
three different modes. It is seen from the enthalpydiagram that in the inductive mode a reasonableamount of kinetic energy is generated (up to 25%).In the capacitive mode, almost all the enthalpy isdue to chemical reactions, i.e., dissociation of theO2 molecules. As can also be seen from Fig. 11, thevibrational energy of the O2 molecules is negligible.To better visualize the vibrational energy, it isplotted together with the chemical mode.A very interesting result is seen in the thermal
mode. Its partition does not follow the sudden in-crease as observed for the other partitions. Thiscan be explained by the mentioned nozzle effect thattakes place inside the discharge tube due to Lorentzforces at high power input [26]. The exhaust velocityfrom a nozzle flow is dependent on the gas tempera-ture T0 in the plenum chamber according to v ∝
ffiffiffiffiT
p.
In an adapted nozzle flow, a high plenum tempera-ture is converted into high velocity throughout thenozzle, while the flow temperature is decreased.The observed increase of kinetic enthalpy at onlyslightly increased temperature between 5000Vand 6000V anode voltage is hence caused by the noz-zle effect. Although plenum temperature may behigh at the point of the measurement, i.e., afterthe magnetic nozzle, thermal energy has alreadybeen converted into kinetic energy. As mentionedin Section 1, assuming thermal and chemical equili-brium, the enthalpy has also been evaluated fromlocal heat flux measurements on copper togetherwith measurements using a calorimeter [2], result-ing in an enthalpy of h ¼ 27MJ=kg. This valueagrees well with the measurements of this paper.Diode laser absorption measurements used byMatsui et al. [6] took into account the frequencyof the plasma. The mean enthalpy they calculatedby averaging over the plasma frequency is h ¼33:7� 3MJ=kg, while the velocity was calculatedin the same way as outlined above. Matsui et al.conclude from their analysis that the local dissocia-tion degree is Δ ≥ 0:92. However, this calculation isbased on the measurements of the excited atomicoxygen.
5. Conclusion
Using quantitative, two-photon absorption laser-induced fluorescence (LIF) measurements of atomicoxygen, the dissociation degree, the translationaltemperatures, and the local mass-specific enthalpyof a pure oxygen plasma flow have been estimated.It can be shown that full dissociation is reachedunder high discharge conditions. For a reference con-dition, data are completed that have been derivedfrom probe measurements at positions farther down-stream. Taking into account the errors due to un-avoidable assumptions for the comparative methodsbased on probe measurements, the data deducedfrom LIF measurments are most reliable. Underthe assumption of thermal equilibrium and neglect-ing ionization, but considering a possible chemicalnonequilibrium, the local mass-specific enthalpy is
derived from experimental data. Nevertheless, thevalues agree well with probe measurements as wellas results from diode laser absorption spectroscopyconducted under this condition, where the influenceof the chemical nonequilibrium effect is rather small.The approach using LIF measurements also offersthe possibility of determining the partitions, i.e. ki-netic, thermal, and chemical partitions, of the localmass-specific enthalpy. An assumed nozzle effect athigh discharge voltages could be strengthened fromthe measured free stream data.
The authors acknowledge the support of H. Böhrk,S. Pidan, and A. Knapp. The authors also gratefullyacknowledge the funding of this work throughDeutsche Forschungsgemeinschaft (DFG) projectAu85/24-1. C. Eichhorn thanks the EvangelischesStudienwerk Villigst for financial and idealistic sup-port through a Ph.D. scholarship.
References1. G. Herdrich, M. Auweter-Kurtz, P. Endlich, and H. Kurtz,
“Mars entry simulation using an inductively heated plasmagenerator,” J. Spacecr. Rockets 40, 690–694 (2003).
2. G. Herdrich, M. Auweter-Kurtz, T. Laux, and M. Winter, “Op-erational behavior of inductively heated plasma source IPG3for entry simulations,” J. Thermophys. Heat Transfer 16, 440–449 (2002).
3. S. Pidan, M. Auweter-Kurtz, G. Herdrich, and M. Fertig, “De-termination of recombination coefficients and spectral emis-sivity of thermal protection materials,” AIAA Paper AIAA-2004-2274 (American Institute of Aeronautics and Astronau-tics, 2004).
4. G. Herdrich, S. Löhle, S. Pidan, M. Auweter-Kurtz, and T.Laux, “IRS ground-testing facilities: thermal protection sys-tem development, code validation and flight experiment de-velopment,” J. Spacecr. Rockets 42, 817–824 (2005).
5. G. Herdrich, “Aufbau Qualifikation und charakterisierungeiner induktiven plasmawindkanalanlage zur simulation at-mosphärischer eintrittsmanöver,” Doctoral dissertation (Uni-versität Stuttgart, Institut für Raumfahrtsysteme, 2004).
6. M. Matsui, K. Komourasaki, G. Herdrich, and M. Auweter-Kurtz, “Enthalpy measurement in inductively heated plasmagenerator flow by laser absorption spectroscopy,” AIAA J. 43,2060–2064 (2005).
7. M. Matsui, K. Komurasaki, G. Herdrich, and M.Auweter-Kurtz, “Temporally resolved measurements of en-thalpy in inductively heated plasma generator flows,” pre-sented at the 4th International Symposium on AtmosphericReentry Vehicles and Systems, Arcachon, France, 21–23March 2005.
8. G. Herdrich, S. Löhle, S. Pidan, M. Auweter-Kurtz, and M.Fertig, Advanced pyrometric in-flight sensor systems for ex-pert, presented at the 4th International Symposium on Atmo-spheric Reentry Vehicles and Systems, Arcachon, France, 21–23 March 2005.
9. S. Löhle and M. Auweter-Kurtz, “Laser-induced fluorescencemeasurements of atomic oxygen using two calibration meth-ods,” J. Thermophys. Heat Transfer 21, 623–629 (2007).
10. S. Löhle, “Untersuchung von wiedereintrittsplasmen mit hilfelaserinduzierter fluoreszenzmessungen”, Doctoral disserta-tion (Universität Stuttgart, Institut für Raumfahrsysteme,2006).
11. K. Niemi, V. S. von der Gathen, and H.-F. Döbele, “Absolutecalibration of atomic density measurements by laser-induced
fluorescence spectroscopy with two-photon excitation,”J. Phys. D 34, 2330–2335 (2001).
12. D. Bamford, L. Jusinski, andW. Bischel, “Absolute two-photonabsorption and three-photon ionization cross sections foratomic oxygen,” Phys. Rev. A 34, 185–198 (1986).
13. A. Goehlich, T. Kawetzki, and H.-F. Döbele, “On absolute cali-bration with xenon of laser diagnostic methods based on two-photon absorption,” J. Chem. Phys. 108, 9362–9370 (1998).
14. W. Wiese, M. Smith, and B. Glennon, Atomic Transition Prob-abilities. Volume I. Hydrogen through Neon, NSRDS-NBS 4(National Bureau of Standards, 1966).
15. H.-F. Döbele, U. Czarnetzki, and A. Goehlich, “Diagnostics ofatoms by laser spectroscopic methods in plasmas and plasma-wall interaction studies (vacuum ultraviolet and two-photontechniques),” Plasma Sources Sci. Technol. 9, 477–491 (2000).
16. H.-F. Döbele, T. Mosbach, K. Niemi, and V. S. von der Gathen,“Laser-induced fluorescencemeasurements of absolute atomicdensities: concepts and limitations,” Plasma Sources Sci.Technol. 14, S31–S41 (2005).
17. H. Griem, Plasma Spectroscopy (McGraw-Hill, 1964).18. G. Abramovich, Theory of Turbulent Jets (MIT Press, 1963).19. W. Vincenti and C. Kruger, Introduction to Physical Gas Dy-
namics (Wiley, 1965).
20. R. Goulard, “On catalytic recombination rates in hyper-sonic stagnation heat transfer,” Jet Propulsion 28, 737–745(1958).
21. C. Park, G. Raiche, D. Driver, J. Olejniczak, I. Terrazas-Salinas, T. Hightower, and T. Sakai, “Comparison of enthalpydetermination methods for arc-jet facility,” J. Thermophys.Heat Transfer 20 672–679 (2006).
22. S. Löhle, M. Auweter-Kurtz, andM. Eberhart, “Local enthalpymeasurements in a supersonic arcjet facility,” J. Thermophys.Heat Transfer 21, 790–795 (2007)..
23. H. Leipmann and A. Roshko, Elements of Gasdynamics(Wiley, 1957).
24. A. Andriatis and I. Sokolowa, “Oxygen. transport andthermodynamic properties.” Mat. Model. 6, 88–128 (1994)(in Russian).
25. B. Bottin, “Aerothermodynamic model of an inductively-coupled plasma wind tunnel”, Doctoral dissertation (VanKarman Institute for Fluid Dynamics, Université de Liège,1999).
26. G. Herdrich, D. Petkow, and M. Auweter-Kurtz, “High enthal-py, water-cooled and thin-walled ICP-sources: characteriza-tion and MHD-optimization,” Cambridge J. Plasma Phys.74, 1–29 (2007).
