next up previous
Next: Conclusions and future research Up: No Title Previous: Comparisons with experiments

Design by optimization

A cavitating blade design technique (CAVOPT-3D) was developed recently [28, 29]. The method consists of a non-linear optimization algorithm which under given certain design conditions and constraints, produces the most efficient cavitating propeller design. Design conditions required by CAVOPT-3D include; advance coefficient ( tex2html_wrap_inline692 ), the wake inflow, cavitation number ( tex2html_wrap_inline694 ), Froude number ( tex2html_wrap_inline696 ), number of blades (NBLADE), hub radius ( tex2html_wrap_inline698 ), and thrust coefficient ( tex2html_wrap_inline612 ). This method also provides the user with the option of imposing constraints on the propeller design. The cavity area constraints, CAMAX and FAMAX, limit the amount of back and face cavity area respectively. The skew constraint, SKMAX, limits the maximum skew angle at the tip of the blade. VVMAX limits the values of the cavity volume velocity harmonics, which are directly related to the magnitude of the hull pressure harmonics. The quantities used in CAVOPT-3D are are determined via quadratic expansions of the results from THPUF-3A in terms of the blade parameters. CAVOPT-3D usually takes several hours on a DEC Alpha to reach a ``convergent'' blade geometry.

The following study uses CAVOPT-3D to produce cavitating propeller geometries with combined forward and backward skew. This is accomplished by allowing for a quadratic expression for skew rather than a linear expression. The description of the skew consists of the skew at the tip of the blade, as well as a new parameter, RFOR, which defines the radius at which the forward skew reaches a maximum value. For this study, the following design conditions were used: tex2html_wrap_inline702 , tex2html_wrap_inline608 , tex2html_wrap_inline706 , NBLADE = 3, tex2html_wrap_inline710 , tex2html_wrap_inline712 . These are the same conditions as given in [28]. The wake inflow is the same to that shown in Fig. 5.

Three cases were run for these design conditions, each with different imposed constraints. A summary of the imposed constraints are given at the top of Figure 11. Run 1 imposes no constraints and no skew. The resulting blade geometry has excessive amounts of cavitation. Run 2 imposes maximum quadratic skew of tex2html_wrap_inline714 (RFOR=0.4) and a maximum cavity to blade area ratio equal to tex2html_wrap_inline718 . Run 3 imposes the same constraints as Run 2 and, in addition, imposes a limit on the cavity volume velocity to a tex2html_wrap_inline720 lower value of that from Run 2. Notice that the resulting geometry slightly violates the constraint on CVV (CVV>VVMAX). Also notice the smaller cavity volume of Run 3 to that of Run 2 (at all blade angles). As expected Run 3 is more loaded at the hub and less loaded at the tip in order to have the cavity volume reduced at the tip while maintaining the same thrust.

  figure180
Figure 11:   Results from CAVOPT-3D. The constraints for each of the three runs are shown at the top part, and the resulting geometries are shown at the bottom part, together with the resulting cavity patterns at a blade angle of tex2html_wrap_inline616 .


next up previous
Next: Conclusions and future research Up: No Title Previous: Comparisons with experiments

Spyros A. Kinnas
Sat Jun 13 10:46:17 CDT 1998