BY F. B. KINGSBURY. (From the Biochemical Laboratories, Department of Physiology, Medical School of the University of Minnesota, Minneapolis.

Transcription

1 A PERFUSION PUMP. BY F. B. KINGSBURY. (From the Biochemical Laboratories, Department of Physiology, Medical School of the University of Minnesota, Minneapolis.) PLATES 2 AND 3. (Received for publication, October 13, 1916.) Several pumps for perfusing surviving organs have been described in the literature, but it seemed to the writer that a new pump of simpler construction could be devised which would be a distinct improvement on those already existing. Two requirements seem to be of fundamental importance: first, that the material of which the pump cylinders, pistons, and valves are made, the material which comes in contact with the blood or other perfusing liquid, should be of glass or hard rubber, and not of metal, since the use of the latter might introduce complications, either because of the possible toxicity of the metal to living material, or because of the chemical action of the metal upon the perfusing liquid; secondly, it seems safe to assume that pulse pressure is of fundamental importance in maintaining the normal functions of any organ or tissue. Gesell,l working with dog kidneys, has recently shown that if the pulse pressure is diminished the functional activity of this organ is distinctly lowered. It seems important, therefore, that a pump to be effective should be capable of imitating the normal pulse pressure curve. Friedmann has devised two rather elaborate perfusion pumps. The material of which the cylinders, pistons, and valves were made appears to the writer to be metal, from the appearance of the photographic reproduction and because of the difficulty of constructing valves such as Friedmann used out of any other material. One of these pumps is so designed, as appears from the drawings, that it would probably deliver a pulse pressure similar to that of an animal. The other pump is not so designed. 1 Gesell, R. A., Am. J. Physiol., 1913, xxxii, Friedmann, E., Biochem. Z., 1910, xxvii,

2 Perfusion Pump The pump described by Richards and Drinkers is a distinct advance in that the cylinders, pistons, and valves are made of glass instead of metal. This pump will produce a pressure curve of the right shape. The chief objection to glass is that it is such a difficult material to work with, as pointed out by these investigators. The valves of this pump are rubber tubes which are mechanically compressed in closing, and properly constructed cams insure the correct timing of the opening and closing. I have found that two Bunsen valves, as shown in Fig. 1, serve this purpose adequately since I have been able to produce pulse pressure curves which are similar in form to the normal arterial pulse curve. These curves are shown in Fig. 2. The t,wo pumps described several years ago by Brodie do not meet the requirements as far as the form of pulse pressure curve is concerned. Hooker has called attention to the fact that a pump in which the crank axle revolves about a fixed point, as is the case with Brodie s pumps, produces a curve having nearly the same velocity of rise and fall, entirely unlike the normal arterial pulse which reaches its maximum quickly to fall off to its minimum slowly. Hooker has made USC of a suitable cam and a sliding pump carriage to attain this end. I assume, from lack of knowledge to the contrary, t,hat the cylinders, pistons, and valves of this pump mere made of metal. Since glass is a difficult material out of which to make cylinders dnd pistons, and because the use of metal for this purpose is inadvisable for the reasons already stated, I have used hard rubber for these parts. Glass valves were first tried, but they proved unsatisfactory, as was also found to be the case by Richards and Drinker, and Bunsen valves were substituted. In this type of valve there is a very slight regurgitation, but this is an advantage over the mechanically operated valves used by other investigators, in the opinion of the writer, since a closer approximation to the normal heart action is attained. Description of the Pump, The pump consists of a hard rubber cylinder snugly fitting an enclosing cylinder of brass, threaded on each end t o receive brass caps. By screwing these up, the hard rubber nozzle of the pump 3 Richards, A. N., and Drinker, C. K., J. Pharm. and Exp. Ther., 1915, vii, Brodie, T. G., J. Physiol., 1903, xxix, 266. Brodie, T. G., and Dixon, W. E., ibid., , xxx, Hool~er, D. R., Am. J. Physiol., , xxvii, 24.

3 F. B. Kingsbury 169 is held in water-tight contact with the cylinder. Text-fig. 1 illustrates the construction and method of attachment of t hese parts. The piston, turned to fit the cylinder, is made of hard rubber and has three grooves in it to receive packing thread. The brass piston rod is connected with a cross head of brass traveling on two steel rods. This is attached to a connecting rod of brass which is in turn clamped by a thumb screw to an oscillating steel beam with a slit in it so that the length of pump stroke may be regulated by clamping the head of the connecting rod at the desired place on the beam. The latter is actuated by a rotating cam on the filling stroke of the pump, and by a steel spring on the emptying stroke. The cam is of the proper shape to produce the characteristic pulse pressure curve, and rotates clockwise. TEXT-FIG inch cylinder and nozzle. Actual size. A, nozzle. B, cylinder. C, ring to adapt nozzle to cylinder (hard rubber). Two sets of pistons and cylinders are provided, one of 3 inch bore and the other of 2 inch bore. By means of the former the output of the pump per stroke may be varied from 0.45 cc. to 2.50 cc., and by the latter from 1.00 cc. to 5.70 cc. The steel beam is marked at certain intervals and the corresponding output volumes ape known. The tension of the spring may be regulated by loosening the lock nut on the beam. shaft and moving the spring arm forward or backward. As shown in Fig. 3, the screw to which the spring is attached is only temporary, the original design calling for a suitable means of adjusting the spring tension at this point also. The spring must be powerful enough to hold the beam against the revolving cam at all periods of the stroke, re- G

4 170 Perfusion Pump gardless of the length of the stroke and the size of cylinder and piston used. In this case the pulse wave is dependent ent irely upon the shape of the cam and is independent of the spring. As shown in Figs. 3 and 4, the cam shaft is equipped with two pulley wheels of different diameters. Either of these may be belted to any one of four grooves in the driving wheel. The latter is attached to a shaft, driven by a worm gear directly connected with a + horse power direct current motor. The motor is started by means of an eight point rheostat, thus giving eight different motor speeds. The worm gear used by the writer gives a speed reduction of 1 to 80, and the possible variations obtainable are from 14 to 96 strokes per minute. Higher speeds may easily be obtained by using a worm gear of smaller speed reduction, and this is recommended, particularly if the organs of small animals, as cats and rabbits, are to be perfused, since the pulse frequency of these animals may be twice or more than twice the maximum frequency obtainable unless a very large driving pulley is used. The worm gear, driving pulleys, motor, and rheostat are not shown in the figures. As previously noted, the pulse pressure curve produced by the pump is dependent only upon the shape of the cam provided the actuating spring is sufficiently heavy. When the pump is adjusted for a large output per stroke, the spring tension must be much greater than when the output is small. If the spring tension is not sufficient to hold the beam against the rotating cam at all positions, a pulse pressure curve is produced which is more or less dependent upon the spring tension itself since the latter does not expel the liquid from the pump with sufficient velocity to hold the beam against the cam as the point of the latter falls away from it. The writer uses several springs of different degrees of stiffness to fulfil these requirements. The amount of power necessary to run the pump is, a priori, greatly increased by increasing the volume of output, and this must be taken into consideration by providing a belt sufficiently heavy to carry the increased load put upon it. Text,-fig. 2 shows the construction of the driving cam. A circle of 2 inch radius is described with C as a center and the diameter ACB drawn. With 0 as a center (OB = 1 inch) an arc BE is described. From C a line drawn at 60 degrees to ACB, intersecting arc BE at P. From P a line is drawn tangent to the circle of which ACB is the diameter, completing the cam.

5 F. B. Kingsbury 171 Text-fig. 1 shows the 2 inch cylinder and its nozzle. The same nozzle fits the 3 inch bore cylinder also. In the assembled pump, shown in Figs. 3 and 4, the nozzle and cylinder are held inwatertight contact by the two brass caps which screw on to the ends of the brass supporting cylinder. It is impossible for the perfusing liquid to come in contact with the brass caps or brass cylinder, or in fact with any meta.llic part. Fig. 1 shows the two cylinders, two pistons, nozzle, and the two Bunsen valves. The latter, simply rubber tubes having short TEXT-FIG. 2. The cam. Actual size. longitudinal slits in them, closed at one end with short pieces of glass rod, are attached to the ends of a Y-tube. This is connected by a rubber tube which is short to lessen the amount of dead space in the pump and connections. Fig. 2 shows kymograph records at different drum speeds of pulse pressure tracings under different conditions of pressure and volume of flow. In making these, a Marcy s tambour with rubber drumhead sufficiently heavy to withstand a high pressure, was connected by means of a T-tube to the outlet valve of the pump. The other end of the T-tube was connected to a

6 172 Perfusion Pump second T-tube, one end of which was attached to a mercury manometer and the other end to a safety valve made by dipping a glass tube to the required depth in a vessel of mercury contained in a large basin. The outflow of the pump then escaped through the column of mercury into the vessel containing it when the required pressure was reached, and flooding this, flowed into the basin. This safety valve also served as a means of producing capillary resistance artificially. The data obtained in making these tracings are shown in Table I. TABLE I. 2 Inch Rare Cylinder. Record A. Record B. Systolic pressure, mm. Hg Diastolic _, Pulse Strokes of pump per minute Flow of fluid per minute, cc The pulse pressures recorded in Table I are somewhat lower than others obtained without connecting in the tambour on account of the fact that the latter, being fairly large and elastic, acted as a reservoir and diminished the difference between the systolic and diastolic pressures. TABLE II. Details of Construction. Brasspumpbase..._..._..._..,. S& x2 x~ standard... 7 high, of A brass. Steel beam.. 62 x 4 x t, 33 slit. I spring arm. _. 32 long. Brass connecting rod., _. 5$ Hard rubber cylinders.. 3& I I pistons.. 1 Brass piston rods.. 3% Maximum stroke of piston... 2 Minimum. $ Cross head standard. _. 3% long, over all. Steel cam shaft. 5; x A beam shaft.. 49 x & Pulleys.. 16 and 1 belt centers. Center of cam shaft to center of beam shaft. 1 horizontal distance. 12 perpendicular distance.

7 F. B. Kingsbury 173 I have used the pump described in the foregoing pages in the perfusion of dog livers and kidneys. It is just as well adapted where a continuous flow is required, in the former case, as in the latter where a pulse pressure is necessary. In the former case the pump pulsations are obliterated by means of an air reservoir in the system. The only parts of the pump requiring much attention are the pistons and these must be freshly packed with thread occasionally. Vaseline is used as lubricant. In concluding I wish to thank Mr. L. U. Boyle, the mechanic of the Medical School of the University of Minnesota, who made the pump, for helpful suggestions in carrying out the details.

8 THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. XXVIII. PLATE 2. FIG. 1. B. FIG. 2. (Kingsbury: Perfusion Pump.)

9 THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. XXVIII. PLATE 3. FIG. 3. FIG. 4. (Kingsbury: Perfusion Pump.)

10 A PERFUSION PUMP F. B. Kingsbury J. Biol. Chem. 1916, 28: Access the most updated version of this article at Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's alerts This article cites 0 references, 0 of which can be accessed free at ml#ref-list-1