2 (0 2 02(g) AH^ HTP - NASA · D:\Documents and Settings\gostorcWy DocumentsWapers 8z Presentations\Welding Effects on €€I’P DecompositionEffect of Welding on Compatibility

!

i

Isothermal Calorimetric Observations of the Effect of Welding on Compatibility of

Stainless Steels with High-Test Hydrogen Peroxide Propellant

Rudy Gostowski, PhD.

Marshall Space Flight Center, National Aeronautical and Space Administration, TD40,

Huntsville, AL

Introduction

High-Test Hydrogen Peroxide (HTP) is receiving renewed interest as a monopropellant

and as the oxidizer for bipropellant systems. HTP is hydrogen peroxide having concentrations

ranging from 70 to 98%. In these applications the energy and oxygen released during

decomposition of HTP is used for propulsion.

2 H202 (0 + 2 H20 (g) + 0 2 ( g ) AH^ = 2887.0 J g-' anhydrous HTP [ 11 (1)

However, incompatibility with structural materials will lead to decomposition of HTP in

areas of the propulsion system that does not result in useful energy production. In addition,

incompatibility may cause undesirable pressure and temperature rises and loss of capacity.

Compatibility is generally thought to be controlled by complex interactions of the surface area,

the chemical constituency and the surface finish of the material. Therefore, implementation of

HTP as a propellant requires testing to determine the compatibility of structural materials for

fabrication of the propulsion system. Compatibility has been expressed functionally with a series

of classes ranging from one to four with one being preferred [2]. Percent Active Oxygen Loss

per Week (%AOUwk) has been used to quantitatively express compatibility.

D:\Documents and Settings\gostorcWy DocumentsWapers & Presentations\Welding Effects on HTP DecornpositionEffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 1 of 7 9/3/03

https://ntrs.nasa.gov/search.jsp?R=20040014946 2018-10-01T06:22:14+00:00Z

%AOUwk= 1 0 0 (Wl C1- W2C2) / W1 C1 (2)

WI and W2 represent the initial and final masses of HTP solution respectively. C1 and C2 are the

initial and final HTP mass fractions. To relate these quantitative measurements to the functional

compatibility designations, metal samples having a %AOUwk of 55 are assigned [3] to Class 1,

while those having a value between 5 and 80 are considered Class 2. Class 3 materials have a

%AOUwk >80.

Isothermal heat-conduction microcalorimetry (IMC) permits measurement of the energy

released when "IF decomposes which indirectly provides %AOUwk data for the sample and

facilitates class ranking.

In propulsion systems components must be fabricated and connected using available

joining processes. Welding is a common joining method for metallic components. The goal of

this study was to compare the HTP compatibility of welded vs. unwelded stainless steel.

Experimental

Propulsion grade high-test hydrogen peroxide (go%, FMC, Inc), reagent hydrogen

peroxide (35%, Spectrum), nitric acid (70%, trace metal grade, Fisher Scientific) and sodium

hydroxide solution (50%, Fisher Scientific) were used as received. Stainless steel sheet (0.15 cm

thickness, CRES 316L and CRES 304L, Penn Stainless Products, Inc.) was cut into panels (10.2

x 10.2 cm). One panel of each alloy was Tungsten Inert Gas (TIG) welded with purge gas

(Helium), one welded without purge gas and one was not welded. The panels were milled into

coupons (3.8 x 1.0 cm). The dimensions of the coupons including the weld bead were

D:V)ocuments and Settings\gostorcWy DocumentsWapers & Presentations\Welding Effects on HTP DecompositionEffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 2 of 7 9/3/03

determined using calipers and the surface area estimated (Table 1). The samples were then

evaluated with a Surface Roughness Tester (Surftest 402 Series 178, Mitutoyo) yielding the root

mean square of the deviations of the surface profile from the mean line (Rq) and the maximum

peak to valley height (Rmax). Results are also listed in Table 1. The coupons were cleaned in

acetone and detergent and rinsed with deionized water. Coupons were passivated by soaking in

70% nitric acid for five hours, rinsed, soaked in 35% hydrogen peroxide for four hours, rinsed in

deionized water and air-dried. Molded borosilicate serum bottles (30m1, Wheaton Science

Products) were passivated by soaking in sodium hydroxide (10%) for one hour, rinsed, soaked in

nitric acid (35%) for one hour, soaked in 35% hydrogen peroxide for twenty four hours, rinsed in

deionized water and air dried.

An Isothermal Microcalorimeter (Model 4400, Water Bath Model 7238, Calorimetry

Sciences Corp, Provo, UT) was used to obtain heat flow measurements at 60 “c. The unit was

permitted to thermally equilibrate at the set temperature for 24 hours and then calibrated against

an internal resistance heater standard. Three measurement cells were evaluated against a

reference cell holding a sealed vial containing deionized water (20mL). Aluminum closures

were used with trifluoroethylene (TFE)-faced silicone liners for the reference and sample vials.

Background heat flows (PB, pW g-’ anhydrous HTP) were measured for three vials containing

only hydrogen peroxide (90%, 20mL). The anhydrous HTP mass was taken as 90% of the total

fluid mass. Heat flows were constant over the 50-hour observation interval.

Vials were opened, metal sample coupons added and recapped. In all cases the heat flow

quickly attained a thermal equilibrium state ( 4 5 hours) followed by a slow rise to a peak value

(Ps+B, p W g-’ anhydrous HTP) within 70 to 320 hours and subsequently diminished. The

D:\Documents and Settings\gostorcWy DocumentsWapers 8z Presentations\Welding Effects on €€I’P DecompositionEffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 3 of 7 9/3/03

background value (PB) representing a sum of the homogeneous HTP decomposition and the

heterogeneous HTP decomposition on the surface of the vial was subtracted from the peak heat

flow (Ps+B). The resulting value (Ps, pW g-' anhydrous HTP) corresponded to the maximum

heterogeneous HTP decomposition on the surface of the coupon.

As shown in Equation 2, PS was converted to a first order rate constant (k, s - I ) by

division with the heat of reaction for the decomposition of hydrogen peroxide (hHr = 2887.0 x

lo6 p W s g -') [l]. Using Equation 3 the resulting first order rate constant was converted from

reciprocal seconds to %AOUwk with the results listed in Table 1.

%AOL/wk = (6.048~10~) k

Results

Table 1

Surface area, finish of samples and percent active oxygen loss per week*

Material Joining Method Area (cm') Surface Roughness Maximum

%AOUwk

Rmax ( P d C Rq ( P I d

CRES 3 16L No weld 10.31M.02 0.97d.25 0 .3M.09 5 .9d .5

CRES 304L No weld

CRES 3 16L TIG, He purge 9.364.14 3.66A.02 1.07M.19 34.h10.0

CRES 304L TIG, He purge 9.434.04 2.29k1.02 1.83k1.11 6 8 . 2 4 1.5

CRES 3 16L TIG, No purge 9.4oLo. 1 1 32.26k14.02 15.42k2.39 36.1k4.1

10.3oJt0.04 1.14M. 1 8 0.33d.15 27.921.6

D :\Documents and Settings\gostorc\My DocumentsWapers & Presentations\Welding Effects on HTP DecompositionEffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 4 of 7 9/3/03

CRES 304L TIG, No purge 9.46j.X). 1 1 106.93~61.21 17.8 k2.32 94.523.1

'Average of three coupons.

%O% confidence level.

'Maximum peak to valley height.

dRmt mean square of the deviations of the surface profile from the mean line.

Unwelded CRES 316L shows good compatibility and by the rating system is nearly Class 1.

CRES 304L is a Class 2 material. Welded CRES 316L and welded CRES 304L have

significantly less compatibility than their unwelded counterparts with the CRES 304L falling

beyond the upward limit of Class 2 materials. When welded without purge gas the CRES 316L

is not significantly different from the metal welded with the gas. However, CRES 304L welded

without purge gas has a much higher %AOUwk and would be considered a Class 3 material.

Discussion

As mentioned previously, compatibility is determined by the surface area, the chemical

constituency and the surface finish of a material. In this investigation exposed area is obviously

not at factor as the welded samples had a slightly smaller surface than the unwelded, but were

more reactive. The chemical makeup of welded CRES 3 16L and welded CRES 304L have been

observed in the literature to change from the parent material as chromium and iron are

segregated in zones [4-61. In particular, the ratio of chromium to iron [6] in CRES 316L

increased from 0.26 to 0.79 in the heat affected zone (HAZ) of the weld and to 1.52 in the weld

bead itself. In CRES 304L the ratio of chromium to iron increased from 0.28 to 0.44 in the HAZ

and to 0.33 in the weld bead. It is possible that the increased reactivity of the welded samples

and of those welded without purge gas is due to this segregation phenomenon. Likewise the

DADocuments and Settings\gostorcUly DocumentsWapers & Presentations\Welding Effects on HTP DecompositionEffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 5 of 7 9/3/03

reactivity increased in keeping with the greater roughness of the welded and welded without

purge gas samples. Therefore enhanced roughness may also be responsible for the increased

reactivity.

Conclusions

In summary, "IF reactivity increased due to welding concurrently with formation of segregation

zones and increased roughness. These effects were even greater when purge gas was not used.

Causality between these factors while reasonable was not established and their fractional

contributions to reactivity were not determined.

Acknowledgments

The author is grateful to Bill Stanton (Marshall Space Flight Center, Metallic Materials Group)

for welding the various specimens and to Yvonne Villegas (NASA-Undergraduate Student

Research Program), Genne Nwosisi (NASA-Minority Programs) and Jennifer Baldridge (NASA-

Undergraduate Student Research Program) for their help in preparation and execution of the

IMC analysis.

References

[ 11 W.C. Schumb, C.N. Satterfield and R.L. Wentworth, Hydrogen Peroxide, Reinhold,

Baltimore, 1955, p. 249.

[2] W.C. Schumb, C.N. Satterfield and R.L. Wentworth, Hydrogen Peroxide, Reinhold,

Baltimore, 1955, p. 163.

D:\Documents and Settings\gostorcUly DocumentsWapers & Presentations\Welding Effects on HTP DecompositionWffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 6 of 7 9/3/03

I

,

[3] Bulletin #104: Materials of Construction for Equipment in Use with Hydrogen Peroxide,

FMC, 1966, p. 48.

[4] T. Takalo, N. Suutala and T. Moisio, Metall. Trans A. 1OA (1979) 1173.

[5] J. Foulds and J. Moteff, Metall. Trans. A. 13A (1982) 173.

[6] M. Ahmad, K.A. Shoaib, M.A. Shaikh and J.I. Akhtar, J. Mat. Sci. 29 (1994) 1169.

D:U)ocuments and Settings\gostorcWy DocumentsWapers & Presentations\Welding Effects on HTP DecompositionEffect of Welding on Compatibility of Stainless Steel with High-Revision 2.rtf Page 7 of 7 9/3/03

I m E

Q) - w O n

d rc

Ql 0 &

4 4 v3

z I

v3 c d * B n

,

6 CA d k

0 - w

I

0- + n b.0 I

0 z- Pl

n ru I

0-

E 0 c.,

0 0 0 0

0 m 3

tm tm

m 3

3 3 I G 0 0 8 &

od) 0 00 0

v)

A4

$--, 0 00 A

A4 s 2 0 4 #

I

v) VI

8 A4 3

0 2

0 3 '1 0 \

4 #

\-

d

r) d

\ n

u"

w 0 0 4

II

4 s

1 a G G

,4'

E

c) 0 k c) ., E

El 0

0 5 Td El 0 0

*+ c,

d) d) 2

U c! m

n

CI, l 0 a, pc

h e

G G E

h

m G zj

v3 cil d)

3 k CI,

d u h e x 0 a 2 E 3 Td 0 rn 'rl

44 0

0 1

c, s E 0

4

0 w I I I

a

0

n

0 0

x 00

E

4

cr; W

3 0 c), 3 0 0 0 F: Td d)

c,

o r (

1 1

a a a a a

21 0

a 3 * w

3 0

Q)

s

I

5

8 I m a,, b o m o m

rn 0

d) k

a! 0

0 I

a E

*+ 8 x ca c 0

CA $4 5 3

Td E a ba U c k

U -d c

d) bQ k 6 ) d ) m U

Tb

d I

m Q) I I u

e e

E 8 5

m n

0 5 Td 0

m c,

&

d) cd 0

+-,

e k 3 0 c!

0

w n

2 8

d) 0 G d) k

3 0 c! a-

h d) F: 0

e 0

G 0 cd c, P x

0

Q 8

8 h n

0-

h c! a a

E" a Tb d) m

lr, m I I I I

O* 0 d=

0

CA $4 3 0

0

d)

ce 2 I CA

a d

CA d) $4

k 5 0 s m d

ba a,

s 2 c,

Y

4 - 5 3 5 a, a

c) a,c 0 0 c 2 5 0 L a,c k ' 5

3 0

m . . 0 , 3 k

h e 0 rn I

c-, cd

0 a, M c4

0 0 0 0 0

ce Tb d

0 c , m

*?=4 k

3 0 4 m

h c! G ce

4

x 0

00 r-:

n

bl)

4 I

00 N II n

4 I m

&? J c, F: a F: 0 0 d) d k k

c, rn

c,

3 3 Q) Q)

0 c, m

I c 0

.l+ +I,

h n

b13 4

I ct 0

h

I k* d) 3

Tb d i tcc

0 Y

0 0 Q)

6) 3 e > . ‘3.

h m . a > 6) . F d m

u c!

Tb G 3 0 0

d k Y I m

k -4 c4

c\ 3 rn I I e I

0 0 0 0 0

0

+I z 2 " 0 0 0 - +I

6 ,

m d- z F ; $ +I

m 0 0 %

+I

o m e m m m +I

" 8 m m T O - d

F \ o \ o

T O - d

I4 \o

m "

$ u

I4 * 0 m

$ u

2 bl)

2 U

A U

O d

2 U

Td a. 0

Td

d) 3 d

U 0 G m -4

3 3 t4 v 0 m

Td 3 cll

\o . d) 4 rn 2 u

3 m y3

2 2 U

u cl * 0 cc, u2

2 u 2 I r! U

'd) 3cr rn W 5

% d) U G 3 0 0

u M

-2 8 d) c,

0 0 0 0

Gl * 0 m

tcc 0 Td c cd

0

Td c cd

c, Td d) 0 rrl

a d)

2 E 0 c

.r(

0 c,

d)

tcc 0 CA

Ifl

$ rj A bl) 3 0 k

8 p! c, cd

bl)

2 c,

A c, *r(

3 bl) d

E u a CA

d) a 4 0

E ctl 0

Td 'dj 3 2

0 cu Gl W

cd a Td

d) 3 4

3 2 c,

tcc 0

2 c,

E ctl 0

4 m Td

d) z 2 x

Td d) CA 1 u

Td c cd

> 4 I

d) w G d) c 0 N

4 2 0 2 c, m

E u 0 c,

s Td CA .r(

c *rl 0 c, Tb

d) c, 2 3 4-J

d) bn k 3 a

0 d) & 2

bl) f? d) cd

m m 0

& 2 a

cd d) rc; 4

CA c, *rl

d) tcc 0 e

0 0

73 c cd

c, c d) -r(

.r(

3 c 0

0 %

w 5 0 A 4-J

Td cd d) 9

E 0

0

cd CA 4

Tb

d)

Tb rj cd

4 3

m I\ 0

Td c cd

c,

2 c, Td * A

3 m m cd E

4 cd 0

Td

d) 3 4

3

E

3 3 0

0

0 c,

2 c,

c o r (

CA

c A w 5 0 k

.r( CA

g a u Tb d)

Td d) 4

3

. .

0 r4 0 *

d- o

CA .r(

.c! rj H 2 d n e e e e e

m E 0

@M m

E 0

m I

4

rn m 0 0 6l v

ce I

2 ce

0

r\ I

4 6 l

0 0

4

m

m 0 a -

0 d- 0 0

0

.

t

0

0

a 5 a

a d I

1

1

a d

a 8 I Ce c 8 I c) 6)

H

0 &

I

'0 n

Ej c) 6)

cil 6) w

I

4 2 z 8 W

3

I

m

E a t bb

v3

0 0

c, F:

s 3 CI,

I Cr) I u

F4 n

d

0

i3 b &

v3 d) U cd s Td cd b 8 Td

5 I

c b, 3

s I 4 2 4

n

d

0

c! 0

E kil

&

2

n

d

0

c! 0

E b &

z W '1 m '1 4

v3 3 z I z 2 8 G

d) G c 0

n

4 2

u F4 d)

'1 v3 d) m Td

Td w * 1 '1

F9 v3 c c) z

e a e e a

aa c)