How to create elliptic paraboloid in Pro/E

How to create elliptic paraboloid if an equation is given
z = x^2/4 + y^2

[Jeff Howard]

How to create elliptic paraboloid
if an equation is given
z = x^2/4 + y^2

Are you asking how to create a curve by equation or wanting the formula for a
specific curve?

Insert / Model Datum / Curve / From Equation, define csys type, etc.

I don't know what the definition of an "elliptic parabaloid" is, but try pasting
the following in as the equation definition after selecting Cartesian coord sys


height = 2
span = 5
x = (span/2) * cos(t*180)
y = (4 * height / span^2) * x^2
Z = x^2/4 + y^2


and see what you get.

[David Janes]

I think you need "t" in there somewhere, say, to get the "span" value from 0 to 5.
But it's generally the engine for generating the curve by actually calculating and
plotting a series of points that the curve passes through.
--
David Janes

"Jeff Howard" <jeff4136@mindspring.com> wrote
How to create elliptic paraboloid
if an equation is given
z = x^2/4 + y^2

Are you asking how to create a curve by equation or wanting the formula for a
specific curve?

Insert / Model Datum / Curve / From Equation, define csys type, etc.

I don't know what the definition of an "elliptic parabaloid" is, but try pasting
the following in as the equation definition after selecting Cartesian coord sys


height = 2
span = 5
x = (span/2) * cos(t*180)
y = (4 * height / span^2) * x^2
Z = x^2/4 + y^2


and see what you get.

Jeff, thanks a lot for your answer. Here is how the Elliptic Paraboloid
looks like
http://mathworld.wolfram.com/EllipticParaboloid.html

What I want to define is a surface not a curve. This surface is defined
thru 2 variables in an equation in form of
z = f(x, y)
In MathCAD software I can easily do it. But I don't know how to do it
in Pro/E

[Doug]

I think a simple solution would be to generate a parabola using an equation
such as

x=4*t
y=x^2
z=0

Then using that parabola to generate a revolved surface.
I was able to generate a shape such as is shown on your link using that
method.

Doug


<mr.xiaoyu.zhou@gmail.com> wrote in message
news:1123160537.652430.79750@g44g2000cwa.googlegro ups.com...

Jeff, thanks a lot for your answer. Here is how the Elliptic Paraboloid
looks like
http://mathworld.wolfram.com/EllipticParaboloid.html

What I want to define is a surface not a curve. This surface is defined
thru 2 variables in an equation in form of
z = f(x, y)
In MathCAD software I can easily do it. But I don't know how to do it
in Pro/E

Doug,Thank you for your input. However, there is basic difference
between your geom and the one on the web. When cutting x-section
parallel to XY plane, your geom gets circle while the web's geom gets
ellipse.

The essence of the problem is whether Pro/E is not able to calculate 2
INDEPENDENT variables at the same time or not. In your equations above,
you made y the dependent variable of x, but y should be independent
variable. I dont't know how PTC customer service will comment on this.
(Our maintenance just expired.)

Xiaoyu

[Jeff Howard]

Also wondering if the surface can be
constructed if two orthograpic sections
are known (variable section sweep,
constant normal, constant section; as in the
attached)

It does appear that the intersection curve shown is truly elliptical, so I think
this may be what you want (?).

[cad/cam/guy]

Is this the same as a ellipsoid , where all curves are parabolic?

You know a pill shaped object..if so then use curves from equations in
all planes
I have all this worked out, but am short of time.. post over at
mcadcentral if this is what you are looking for

[cad/cam/guy]

this is cut in half otherwise it is full and does become a solid with
no seam

http://www.flickr.com/photo_zoom.gne?id=31633615&size=m

full and radius edged hole solid

http://www.flickr.com/photo_zoom.gne?id=31634904&size=o

Thank you all for your help. cad/cam/guy, would you let us know how you
create it or send file to my email
mr.xiaoyu.zhou@gmail.com
Thanks again

[cad/cam/guy]

first I create 3 seperate curve to use for height,major,minor dim ...
change the dims to names, so you can modify the shape without having to
change the values in the equations
then create some additional cys for the formulas.... you could just
change the xyz values in the eqs. but I just made cys to allow pasting
the equations;

/* --------------------xy_drive_curve --------------------
ratio=minor/major
x=major*cos(t*360)
y=major*sin(t*360)*ratio
z=0
/*--------------------yz_sect_curve1 --------------------
ratio=major/ht
x=ht*cos(t*180)
y=ht*sin(t*180)*ratio
z=0
/*---------------------yz_sect_curve2 ------------------
ratio=major/ht
x=-ht*cos(t*180)
y=-ht*sin(t*180)*ratio
z=0
/*------------------- xz_sect_curve1 -------------------
ratio=minor/ht
x=ht*cos(t*180)
y=ht*sin(t*180)*ratio
z=0
/*------------------ xz_sect_curve2 -------------------
ratio=minor/ht
x=-ht*cos(t*180)
y=-ht*sin(t*180)*ratio
z=0

then surfrace by boundries.. with the xy curve being curve 1.. the
others being 2nd direction... but in reality after thinking about
this..it's not really a parabola... a parabola will not close
up....it's close.. but not true.. anyway I am going to play with this
doing some curve>intersection of some cones.. see where that leads

[Jeff Howard]

... but in reality after thinking about
this..it's not really a parabola... a parabola
will not close up....it's close.. but not true..
anyway I am going to play with this doing some
curve intersection of some cones.. see where that leads

I think it depends on "how close gets the cigar", though I don't pretend to
understand the shape (or what Mr X is after in the grand scheme).

The problem is going to be with conic definitions:

Degree two curves with three CV's (conic arcs)...

Ellipse: 0 < rho < 0.5 (for a quadrant it's always 0.414,
increasing rho will push the point of max curvature away
from quadrant point)

Parabola: rho = 0.5 (always, tangent angles define the "portion"
of curve defined)

Hyperbola: 0.5 < rho < 1

--------------------------------

*If* what I did with the VSS is close to what's desired and a closed shape such
as the ellipsoid is desired I think I'd planar trim the shape, mirror it, and do
a G2 boundary blend between the two. The "long axis normal" xsec will be an
ellipse. The position of the trim plane will define the max girth. A little
layout geometry could be used to define the trim for any given max girth.

Interesting stuff ...

(Speaking of which: CCG, if I haven't confused aliases; I did get your message a
while back. SHTF about that time and I forgot about it. If you are interested
in an example of the cylindrical cam groove, holler and I can email it.)

--------------------------------