^liiE INERTIA OF THE CONNECTING ROD. — MACGREG6R. ^61 



the unit of force in terms of which the scale of the force curve 

 will be expressed will be the unit of force of the system derived 

 from the unit of mass selected and the units of length and time 

 employed above. Thus, the scale of the ordinates of ff, consid- 

 ered as an acceleration curve, having been found, with the 



velocity and distance scales originally selected, to be ft - 



per-sec per sec. to a division, if the mass of the connecting rod 

 be 1 cwt. (British) and if we select the pound as unit of mass, 



the force scale will be 112 x —^ poundals to a division. If we 



wish the force scale to be expressed in pounds-weight, we must 



express ni in terms of the unit of mass of the gravitational system, 



112 400 

 viz., 32 lbs., in which case the force scale is — "^ X — pounds- 



32 3 ^ 



weight to a division. 



The S' — 8 curve having thus been obtained and its scale deter- 

 mined, it is easy to obtain the value of S when P is given, by a 

 graphical method. For, as seen above, 



P(pls)^S', 

 and P being given for all crank positions, and p and 8 being 

 found from the diagrams similar to Fig. 1, the values of 

 Pfp/s) may be found by the ordinary graphical methods of mul- 

 tiplication and division, and a curve of P(jjjs) or S' plotted in the 

 same way as the above curves. Then the excess of the length 

 of any ordinate of the S' curve over that of the corresponding 

 ordinate of the >S' — <S curve, proper regard being had to sign, 

 will be the length of the corresponding ordinate of the S curve. 



Sometimes instead of requiring to find the amount by which 

 the effort on the crank-pin, S, is diminished by the inertia of the 

 connecting rod, the " nett forward piston pressure," P, being given, 

 we have to find the amount by which this latter force must be 

 increased in order tliat the eftbrt on the crank-pin may not be 

 diminished by the inertia of the rod. To make this determina- 

 tion additional calculation is necessary. For we have as before 



s dc 



