Can be done - but I need more data:
pressure angle
major dia
minor dia
diametral pitch
circular tooth thickness (at pitch dia)
gearguru

As gearguru states
you must submit the diametral pitch
or more information.  Splines are often
30 degree involute pressure angles and like
double stub addendums and dedendums.
Any gear program can draw these for you
if the id and od are given.  I assume these
are side fit applications.  The diametral
pitch is normally something like 8/16
or some combinations of numbers.

G'day,
Look on page 11 of this link for spline info:
http://www.pdfx-mcg.com/pdfx_denisonhydraulics/pdfs/PV-PVT-Sales.pdf


NUMBER OF TEETH    13
PITCH              16/32
PRESSURE ANGLE     30°
MAJOR DIA.         0.875/0.853
MINOR DIA.         0.715
SAE 22-4 (SAE B) INVOLUTE SPLINE DATA
CLASS 5, FLAT ROOT SIDE FIT PER J498c

BR
Ian

The information posted by Deaks is correct for the output shaft of my pump.  

     I believe I have a good drawing of the pump spline on my computer now.  I spent a couple hours searching this forum and came upon a math formula for splines.

x = r(cos(theta) + (theta)sin(theta))
y = r(sin(theta) - (theta)cos(theta))
where (theta) is in radians

     Another spec sheet that I got from denison listed the max tooth diameter of of .0982".  

     Using AutoCAD I drew the pitch circle.  Then I drew a rack tooth with 30degree angles over top of the pitch circle.  I moved the rack tooth so that the two angles of the rack teeth intersected the pitch circle, and the pitch circle created an arc from one tooth to the next of .0982".

     I used excel to create a quick spread sheet of x and y coordinates vs (theta).  I figured that (theta) from 0 to 30 degrees with a step size of 3 degrees would be sufficient.  So I had 10 (theta)s, 10 x coordinates, and 10 y coordinates.

     I then copied the rack tooth, rotated it 3 degrees, and moved it to the x and y coordinate that corresponded to 3 degrees of rotation.  Then I copied and rotated the origional tooth 6 degrees, moved it to the x and y coordinates pertaining to 6 degrees etc.

     After I was done I had a tooth profile that looked pretty good.  I had to trim all of the lines that didn't touch the inside profile until I ended up with the outline of the gear tooth.  I also trimed any lines which extended beyond the minimum and maximum diameter for the spline. Afterwords I arrayed the spline around the pitch circle 12 times, for 13 teeth.

    Next, I did the true test to see if the spline was the correct size.  I printed the picture one to one scale, cut the spline out with an exacto knife, and slid the paper onto the shaft.  It was a good fit.

   Now I'm going to have a blank part cut out of scrap metal and see if it fits on the shaft.  Then I can get the desired interference fit I am looking for.

     I know its hard to explain all of this with words, but does it seem like I did it right?  Was there an easier way with autoCAD?  Let me know what you think, I can send my AutoCAD drawings if you like.  (I saved the drawing several times during the process, to document how I made the spline.)


Thank You
Greg Gill

I have an autolisp program that
draws gears and the 30 degree spline
is just a special gear.

It is as follows:

;GEAR.LSP - This program generates involute curve profile.
;Program by Jim Geisey 10-14-90
;USER INPUTS ARE:
;   Number of Teeth     Diametral Pitch
;   Pressure Angle      Arc Tooth Thickness (TT x DP)
;   Addendum            Dedendum
;   Fillet Radius
(defun c:gear ()
  (setq numt (getint "Number of Teeth:<24> "))
  (if (= numt nil)
    (setq numt 24))
    (setq diap (getreal "Diametral Pitch:<1.00> "))
  (if (= diap nil)
    (setq diap 1.000))
    (setq prsa (getreal "Pressure Angle:<20.0> "))
  (if (= prsa nil)
    (setq prsa 20.0))
    (setq adum (getreal "Addendum:<1.0000> "))
  (if (= adum nil)
    (setq adum 1.0000))
    (setq ddum (getreal "Dedendum:<1.1571> "))
  (if (= ddum nil)
    (setq ddum 1.1571))
    (setq tt (getreal "Arc Tooth Thickness :<1.5708> "))
  (if (= tt nil)
    (setq tt 1.5708))
    (setq filr (getreal "Fillet Radius:<0.235> "))
  (if (= filr nil)
    (setq filr 0.235))
    (setvar "orthomode" 0)
    (setvar "coords" 1)
;START CALCULATIONS
;
  (setq pnts 19) ;determines involute curve accuracy
  (setq prsa (/ (* prsa pi) 180.0)) ;pressure angle to radians
  (setq pitr (/ numt (* diap 2.0))) ;calc pitch radius
  (setq irad (- pitr ddum))         ;calc inside radius
  (setq orad (+ pitr adum))         ;calc outer radius
  (setq basr (* pitr (cos prsa)))   ;calc base radius of gear
  (setq z (- (expt orad 2.0) (expt basr 2.0)))  ;three lines of
  (setq x (sqrt z))                             ;code to make
  (setq paodd (atan (/ x basr)))                ;an arccosine
  (setq incr (/ paodd pnts))
  (setq p 0.0)
  (setq tt (* diap (/ tt numt)))
  (GRAPHSCR)
  (setq p2 '(0 0))     ;Center of Gear
  (setq x2 (CADR p2))
  (setq y2 (CAR p2))
  (setq y5 (+ y2 basr))
  (setq y6 (+ y2 irad))
  (setq p6 (list x2 y6))
  (setq p5 (list x2 y5))
  (setq pz1 (polar p2 1.95 (* basr 0.9)))
  (setq pz2 (polar p2 1.35 (* orad 1.5)))
   (COMMAND "CIRCLE" P2 pitr) ;place pitch circle
   (COMMAND "CIRCLE" P2 basr) ;place base circle
   (COMMAND "CIRCLE" P2 irad) ;place inside circle
  (COMMAND "ZOOM" "w" pz1 pz2) ;Window for drawing curve
  (setq s (ssadd))
  (setq le (entlast))
;
;begin a loop for placement of coordinate pairs
;the command pline will be used to generate a polyline whose
;vertices will be computed by a standard involute curve formula
;
  (setq test 0)
  (COMMAND "PLINE" p5)
  (setq p (+ incr p))
  (while (> pnts 0)
    (setq e1 (sin p))
    (setq e2 (cos p))
    (setq e (/ e1 e2))
    (setq j (- e p))
    (setq x1 (* (/ (sin j) (cos p)) basr))
    (setq y1 (* (/ (cos j) (cos p)) basr))
    (setq x3 (+ x2 x1))
    (setq y3 (+ y2 y1))
    (setq p3 (list x3 y3))
    (COMMAND p3)
    (setq p (+ incr p))
    (setq pnts (- pnts 1))
    (if (/= test 1)
      (PROGN
        (setq hyp (sqrt (+ (expt x1 2) (expt y1 2))))
        (if (> hyp irad)
          (progn
            (setq pint p3)
            (setq test 1)))))
    )
  (COMMAND "")
  (initget "Y y N n")
  (setq ans Y)
  (if (/= ans "N")
    (PROGN
      (setq pz3 (polar p2 (* pi 1.25) (* orad 1.6)))
      (setq pz4 (polar p2 (* pi 0.25) (* orad 1.6)))
 ;    (COMMAND "ZOOM" "W" pz3 pz4)
      (setq ang13 (/ (sin prsa) (cos prsa)))
      (setq ang11 (- ang13 prsa))
      (setq x11 (* (sin ang11) pitr))
      (setq y11 (* (cos ang11) pitr))
      (setq p11 (list x11 y11))
      (setq ang (angle p2 p11))
      (setq angi (- ang tt))
      (setq p12 (polar p2 angi (/ ORAD 0.8)))
      (setq beta (angle p2 p3))
      (setq ang2 (- (* angi 2.0) beta))
      (setq p15 (polar p2 ang2 orad))
      (setq adj1 (- ang2 j))
      (setq adj2 (+ adj1 (/ (* pi 2.0) numt)))
      (setq adj3 (+ adj2 0.3))
      (setq p22 (polar p2 adj3 basr))
      (setq p17 (polar p2 adj2 irad))
      (setq p18 (polar p2 adj2 basr))
      (setq angr (* (+ ang11 tt) (/ 180.0 pi)))
      (setq p21 (polar p2 ang2 (/ orad 0.8)))
      (setq angm (+ angi (/ pi numt)))
      (setq p19 (polar p2 angm irad))
      (setq p20 (polar p2 angm (+ irad filr)))
      (COMMAND "ARC" p15 "c" p2 p3)
;     (COMMAND "ARC" p19 "E" P5 "R" filr)
;     (COMMAND "ARC" p6 "c" p2 p17)
;    (COMMAND "LINE" p2 p12 "")
;    (COMMAND "LINE" p2 p20 "")
      (COMMAND "MIRROR" "W" p22 p21 "" p2 p12 "")
 ;   (COMMAND "MIRROR" "W" P18 P21 "" P2 P12 "")
      (COMMAND "ROTATE" "w" pz3 pz4 "" p2 angr)
      (while (setq le (entnext le))
        (ssadd le s)
      )
      (COMMAND "ARRAY" s "" "p" p2 numt "" "")
 )  )
  (Prompt "\nDone!  ")
  (princ)
)
 
Have fun,
Remember that an internal spline is
a mirror image of an external spline if
you reverse the addendums and dedendums
and enter the tooth space instead of the
tooth thickness.  
The fillet radius portion
is not working and you will have to trim
and fillet the involute profiles yourself.
It does not matter what value you enter for
the fillet radius as it presently is not
used in the program.

Greg,
It would be very unusual for someone to use a coupling with an interference fit on a hydraulic pump input shaft, as usually internal damage would result on fitting this type of coupling. Are you doing something exotic?
BR
Ian