M . W. KIEBLER
C oal R e s e a r c h L a b o r a to r yC a rn eg ie I n s t i t u t e o f T e c h n o lo g y , P itts b u r g h , P a.
T
H E use of pressure equipm ent in research laboratories has increased steadily over th e last several decades to th e point of being com m onplace. M an y investigators have found it possible, th ro u g h tem p eratu re or m ass action effects, to increase g reatly a reaction velocity or to shift an equilibrium in a desirable direction b y conducting th e reaction under pressure. In those cases w here th e reactan ts occur in a single phase, such as gaseous or liquid solutions, statio n ary ap p aratu s is often satis
factory. H ow ever, if tw o or even th ree phases are involved, th e ra te of reaction can be fu rth er increased by providing some m eans of agitation. A t th e sam e tim e h e a t transfer betw een th e stirred m ass an d th e vessel wall is m aterially increased. N um er
ous devices to m ix or s tir th e contents of a pressure vessel have been rep o rted in th e literatu re. Phillips (IS) and Groggins and H ellbaek (7) described an arrangem ent whereby cylindrical bom bs a re ro ta te d in a co n stan t tem p eratu re b ath . O ther de
signs which depend on either ro tatio n of th e reactor itself or of in te rn a l blades for ag itatio n are described b y Tongue (IS), L aupichler (9), a n d Fischer (5). Several authors ( i, 4 ,1 1 ) have described a sh aking or rocking autoclave of a ty p e which has been com m ercially available for a num ber of years.
M echanical difficulties—such as bleeding gases in a n d /o r out of a vessel, a tta ch in g gages, condensers, h e a t exchangers, or safety disks, m aintaining co n stan t pressure, etc.—which are encountered w ith ro tatin g , shaking, or rocking autoclaves can be g reatly reduced b y an externally powered and internally stirred autoclave (9, IS ). H ow ever, th is ty p e of equipm ent is n ot en tirely w ith o u t fau lt. I f th e vessel is to be used a t pressures in excess of several th o u san d pounds, it is difficult to obtain a stuffing box w hich will rem ain g astig h t for any considerable len g th of tim e; an d while gas leakage can be reduced by a lan tern -lu b ricated gland, contam ination of th e reaction m ixture w ith th e lu b ric a n t th e n becomes a problem . Pow er losses, th ro u g h a packing gland, resu lt in a high in itial and operating cost an d ten d to reduce th e ro tatio n speed of th e stirrer so th a t, from th e sta n d p o in t of ag itatio n an d cost, th e value of this ty p e of autoclave is often questionable. A 750-cc. autoclave b u ilt in
th e Coal R esearch L ab o rato ry for service a t a p ressure of 3000 to 6000 pounds per square inch req u ired 1/ i horsepow er to tu rn a stirrer a t 120 revolutions p er m inute. T ongue ( I S ) described several sm all lab o rato ry autoclaves of th e sam e ty p e w hich had sim ilar pow er requirem ents. T hese fau lts exceed reasonable limits as th e size of th e au to clav e is reduced, since am o u n t of leakage, size of power in stallatio n , a n d a m o u n t of a g itatio n will rem ain fairly co n stan t w ith wide v a ria tio n in th e volum e of th e pressure vessel.
A m ore nearly ideal re ac to r w ould b e one in w hich th e stirring mechanism an d re ac ta n ts are contained w ith in th e sam e wall.
C alvert (S) in 1914 o b tain ed a p a te n t on th is basic id ea which covered n o t only m otor-driven stirrers, b u t also circulating pum ps. M acM illan an d K rase (10) a n d H ollow ay (8) published a detailed description of an a u to clav e in w hich th e s tirre r an d m otor operate u nder th e sam e gas pressure. R ecen tly (6) a m ethod of obtaining a g ita tio n b y m eans of a m agnetically oper
ate d plunger was reported.
T h is p ap er describes a to ta lly enclosed m o to r an d ag itato r, b u ilt to o p erate u n d er m ore severe conditions w ith respect to pressure, tem p eratu re, a n d chem ical a tta c k th a n those m en
tioned above. T h e stirrer assem bly (F igure 1) is con stru cted as an integral u n it w hich is a tta c h e d to th e bom b h e a d b y m eans of th e th rea d ed lower end. T h e autoclave, show n dism an tled as well as assem bled to ru n in F igure 2, was b u ilt in th ese labora
tories several years ago to s tu d y th e h y drogenation of coal in aqueous alkali a t tem p eratu res an d pressures u p to 400° C.
750° F .) a n d 6000 pounds per sq u are inch. V iolent a g ita tio n is required in th is reaction to produce n o t only th e m axim um possible gas-liquid interface, b u t to p re v en t th e coal p articles from fusing together. Sufficient tu rb u len ce w as o b tain ed in th is auto clav e b y ro tatin g a 2-inch nickel propeller a t 1500 r.p.m . in a 750-cc.
nickel-lined cylindrical bom b of 3-inch in te rn a l diam eter.
T h e body an d to p closure for th e assem bly were m achined from chrom e-vanadium steel (SAE-6145) forgings w hich w ere h e at-trea te d an d draw n a t 900° F . in a s a lt b a th after all m achine w ork h a d been finished. D a ta available for th is alloy in d icate
I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y th a t th e resulting p roduct should have a tensile strength of
210,000 pounds per square inch an d a Brinell hardness of 470 (48 Rockwell C ). T he stu d s were m ade of a chrome-nickel steel (SAE-3140) draw n a t slightly lower tem peratures to give a ten
sile stren g th of ab o u t 200,000 pounds per square inch. The nuts, m ade of th e sam e alloy except for a lower carbon content (SAE-3120), were carburized for surface hardness. T o reduce contam ination of th e bom b contents, th e interior of the assembly was given a thick coat of silver and burnished. T he top closure was m ade up ag ain st an unconfined copper gasket which had an internal diam eter of 4 inches. Silver gaskets were used between the assembly and head an d betw een th e head and th e reactor body. A tig h t-fittin g copper condenser on th e neck of the assembly served to remove th e h e at which flowed up from the heated reactor. Six nickel disks were attach ed to th e nickel stirrer sh aft to create turbulence and expedite h eat removal.
All p a rts of th e electrical system had to be resistan t to a tta ck from hydrogen, w ater, an d any organic vapors which m ight be formed during th e course of a reaction. A shaded-pole 1/30- horsepower m otor was obtained (model 123, A. G. R edm ond Company), and th e ou ter cover removed. Two bearing supports (Figure 1) were m ade for th e top and bottom of th e m otor. An Oilite bearing was pressed in to each support to tak e th e rotor shaft. I t was planned to replace these bearings w ith graphite bushings if hydrogen or organic vapors should remove th e lubri
cant, b ut this has never been necessary. T he supports were held tightly against th e sta to r ends by tw o b en t bolts which passed through slots a t opposite sides of th e stato r. These bolts were salvaged from th e m otor parts.
Two stato rs have given satisfactory service in this apparatus.
The first was tak en as it came from th e m anufacturer and was given about six coats of a B akelite varnish. T he second was made by winding Form ex wire onto a Redm ond sta to r frame and alternately coating th e coils w ith H eresite varnish (No. L-100) and General E lectric varnish No. 1676. Each sta to r was im
pregnated by suspending it in a container of varnish which could be placed in a vacuum desiccator. T he pressure in th e desiccator was then alternately decreased and allowed to retu rn to atm os
pheric to remove entrapped air. T he sta to r was then drained, allowed to air-dry, an d finally baked. A fter each of several im
pregnations, th e s ta to r was supported in a different position for drying and baking. Any varnish which adhered to the outside diameter of th e s ta to r was removed w ith sandpaper. T he as
sembled m otor was tested b y allowing it to run under w ater.
Because this m otor has no sta rte r brushes or windings, only one lead has to be taken o ut through th e reactor wall with, the other lead internally grounded. However, in this design both leads were brought o ut so th a t th e m otor insulation could be more accurately checked. T he leads were made by brazing a copper disk on a short piece of heavy Nichrome wire and slipping a Bakelite disk over th e wire on each side. T he three disks were then crushed together u nder a gland n u t in th e top closure.
Both the diam eter an d thickness of th e Bakelite m ust not be less than th a t of th e copper disk. N ichrom e wire was
used in preference to copper because of greater strength and stiffness. T he resistance to ground through this type of seal (#) has alw ays been more than 500,000 ohms.
A coil which serves to rem ove h e at from the motor sta to r was m ade b y bending a piece of lA X */i inch mild steel tubing to form a long narrow U, which was then wound in to a helix. T he two ends were separated 180° a p art, b en t a t right angles to th e helix, and dropped through holes a t the bottom of the m otor com partm ent w here they were made up against th e wall w ith compression cones. The only change contem plated for this autoclave is to increase th e internal diam eter of
tubing used in the coil. Be a r in g Su p p o r t F igu re 1. B om b Stirrer A ssem bly
340 I N D U S T R I A L A N D E N G I N E E R I N G Vol. 37, No. 6
T h e final assem bly was m ade by attach in g th e stirrer sh aft to th e m otor b y m eans of a bushing, an d th en pushing th e sh aft and m o to r th rough th e coil so th a t th e m otor rested against a shoulder n e ar th e lower p a rt of th e com partm ent. A ring which served to
held a g rap h ite bushing to ta k e th e s tirre r sh aft. A t v a rio u s tim es th is fittin g has been replaced w ith a sim ple packing gla» * in which case th e pressure in th e assem bly a n d bom b were ec^ . ~ ized b y a connecting tube. T h e assem bly a n d head, whic
F ig u re 2. O pen A u to cla v e a t L eft, S h o w in g H ead C o n str u c tio n , T h e r m o c o u p le T u b e, Stirrer, a nd B earing; A ssem b led A u to cla v e a t R ig h t
c en ter th e m o to r an d hold it firmly against th e shoulder was then pressed ag ain st th e top bearing support, and a low-melting alloy (20% lead -4 0 % tin -4 0 % bism uth, m elting a t 202° F.) was poured th rough a hole in th e ring. T his ring, which has two set screws to hold it in place, was rem oved after th e alloy had solidi
fied to facilitate m aking u p th e electric connections. T he alloy provides good h e a t transfer betw een th e coil and stator. The m o to r can be rem oved b y tu rn in g th e assembly upside down and passing superheated steam through th e coil to m elt the alloy.
T he g asket was p u t in place, several stu d s were removed, and th e to p closure was held a sh o rt distance above th e assembly so th a t th e m o to r leads could be soldered to th e N ichrom e wires. After all exposed wires were carefully insulated and coated with thick
ened varnish, th e final closure was m ade.
T h e solid nickel head to w hich th e assem bly was attach ed was fitte d w ith a therm ocouple well, a safety disk, a gas inlet, and an o u tle t w hich could be used in conjunction w ith a siphon tu b e to charge a n d em p ty th e bom b w ith o u t rem oving th e head. T he nickel fittin g on th e lower side of th e head, shown in Figure 2,
weighed ab o u t 80 pounds, could be conveniently raised above th e statio n ary nickel-lined bom b by a cable an d b y one fixed a n d one m ovable pulley.
LIT E R A T U R E C IT E D
(1) A d k in s, H ., In d. En g. Ch e m., An a l. Ed., 4, 3 4 2 -5 , 3 7 9 (1932) (2) Asbury, R. S., Ibid., 8, 152 (1936).
(3) Calvert, G., U. B. Patent 1,123,092 (1914).
(4) Dykstra, F. J., and Calingaert, G., In d. En g. Ch e m An a l Ed., 6, 3 8 3 - 4 (1934).
(5) Fischer, F., G«i. Abhandl. Kermtnw KoUe, 4, 13-25 (19 1 9 ).
(6) Gilson, A . R., and Baskerville, T. W., Chemistry £ Industry 63
4 5 0 (1943). '
(7) Groggina, P. H., and Hellbaok, R., Chem. & Met. Eng 37 693-4 (1930).
(8) Holloway, J, H., and Krase, N. W., In d. En g. Chem 2 5 4 9 7
-502 (19 3 3 ). ’’ ’
(9) Laupiehler, F., Chem. Fabrik, 5, 3 0 5 -1 1 (1932).
(10) MacMillan, A . H., and Krase, N. W., In d. Eno Chim im 1 0 0 1 -2 (1932).
(11) Peters, F. N., Jr., and Stanger, O. C., Ibid., 20, 74-6 (19281 (12) Phillips, M„ Ibid., 17, 721-5 (1925). v >' (13) Tongue, H., “Design and Construction of High p rBmil„
Chemical Plants”, London, Chapman <fc Hall, 1 9 3 4,