Geometry help, please?

sure I agree with that 100%. check it based on the actual tank.

but it would be cool to make a test curve based on what we came up with and then compare to see how close it is :)

you also have to worry that the original measurements were taken properly and are accurately. if the measurements were off at all, it would make the math bad.
 
Last edited:
LOL, first of all, thank you guys so much! I was a little shocked to see 20 replies to a thread I was a little embarrassed to start in the first place. I've gotten more than I could ever have asked, you guys rock.
Anyways, the deal is, I'm a super senior in college on spring break, and the canopy is for my tank at school and I'm at home. I didn't quite have the foresight to just trace the footprint of the tank, I took measurements instead. This worked out great for the rectangular tank, but is proving difficult with the bowfront.

Basically I'm going to cut a triangle for the triangular section of the tank, and cut an arched piece for the bowfront section by hand, but I don't know the arc to cut; that's what I am trying to figure out.
I have the dimensions I took drawn out in scale on a large board. I'd be thrilled to test the numbers you guys came up with, but can I do it without having the tank with me?
I did go 28.1" back from the center of the arc and pivoted in a circular motion and it did hit either end of the sides almost exactly... not sure if that tells you anything.

RyanBrucks, you may get your wish :) I am pretty confident in my measurements, and I am willing to just take it on faith, use the numbers you guys came up with and build it and see it how it fits. I have nothing else to do, lol.
 
Another option is to find a tank locally...maybe a local club member or a LFS has one that you could make a template from?

Could try calling the manufacturer and asking them for the details of the curve...?

Google up "36g corner aquarium" first thing you will see is a thread on another forum where I guy built a stand for his tank...maybe send him a PM...
 
<a href=showthread.php?s=&postid=12025775#post12025775 target=_blank>Originally posted</a> by CoreyK
I have the dimensions I took drawn out in scale on a large board. I'd be thrilled to test the numbers you guys came up with, but can I do it without having the tank with me?
I did go 28.1" back from the center of the arc and pivoted in a circular motion and it did hit either end of the sides almost exactly... not sure if that tells you anything.

What you did just proves that Ryan's math is right. You'll have to have the tank in front of you to make sure that the bow actually has that arc, and not some random non-circular arc. I would bet that the arc is circular.

Welcome to RC!
 
If the front of the tank is an arc of a circle, them the 3 points shown define the ONLY circle that can intersect thos 3 points. Straight forward :)
 
or i guess you're saying that you could just guess-n-check until you find a point from where the radius touches the arc on all points? sounds like thats what he did when he said 28.1 touched both ends.
 
No you don't need to guess.

The center of the circle is a point (the only point) that is equidistant from the 3 points.

We know the equation for a circle with 3 points and we know three points on this circle. How? You guys already did the work.

You know the length of the hypotenuse. To make it easy turn the right triangle on end so that the hypotenuse is vertical and the arc is to the right. set the coordinates of the:

bottom point to 0,0
The top point to 0,(hypot)
the arc point to 1/2(hypot), 7

You have they xy coords for all 3 points.

We know the legs of the right triangle are 26.25 and using a^2+b^2=c^2 gives us the length of the hypotenuse.

So 26.25^2 + 26.26^2 = hypot ^2
or 37.123

So [0,0] [0,37.123] [7,18.56]

Now we can use the equation
(x^2+y^2-x3^2-y3^2)*(x1-x3)*(y2-y3)
+ (x1^2+y1^2-x3^2-y3^2)*(x2-x3)*(y-y3)
+ (x2^2+y2^2-x3^2-y3^2)*(x-x3)*(y1-y3)
- (x^2+y^2-x3^2-y3^2)*(x2-x3)*(y1-y3)
- (x1^2+y1^2-x3^2-y3^2)*(x-x3)*(y2-y3)
- (x2^2+y2^2-x3^2-y3^2)*(x1-x3)*(y-y3) = 0

This equation is satisfied by all points whose coordinates are (x,y)
on the circle.

Expand all that out, and gather together all terms involving x^2, x,
y^2, and y. Then complete the square on x and y to put the equation in
the form

(x-h)^2 + (y-k)^2 = r^2.

Then the center of the circle is (h,k) and the radius is r.

I am too lazy to do the math...

See
http://mathforum.org/library/drmath/view/55166.html for the formula given above. There are several other ways to go about this. You can also figure out the perpendicular bisectors, use determinants, use triangles... etc.

have fun. If it were me, I would go find the same tank and make a pattern!
 
( A rescue attempt) ;)

saltyseaman :) Stud-

How have you been my friend? My first trades with you were grand! Excellent size and color; thanks for giving me my virgin SPS fix.

I followed every post of your bizarre scuba trip; it lured me to the SPS side. Thank you.

How's life been treating you, golden i hope? School and Bride doing OK? Drop me a note.

Sorry, now back to the 101 Trig class.

Chaz
 
Last edited:
ah ok I remember that equation vaguely now. but my algebra skills are too atrophied to do all the expand and gather with that many terms :) or at least it would take me a while hehehe.

assuming that the radius is 28.11, cant you just trace back from the arc midpoint to find the center (using the tank corner as a guideline to pass through)? from the tank corner to the arc midpoint is 25.56, so the center should just be 2.55" away from the tank corner (along the axis from corner to arc midpoint).

whats the easiest way to scribe a circle of that size? I've actually never scribed a circle larger than what a handheld compass could do. I know you could probably do some homemade rig with a piece of string and a pencil or something, but is whats the name for the tool thats meant for that kinda thing?
 
nice to see you Charles!

Ryan:
The problem is that you don't know the diameter. We know the length of the chord (because we know the angle of the arc given the right triangle) and the height of the perpendicular to the chord.

We can also solve with this: (much easier, but what fun is that!)
radius = (m² + 1/4c²)/2m

Where:
m = the height of the perpendicular to the chord (at the midpoint)
c= the length of the chord

So radius = (7^2 + .25(37.123^2)/(2*7) = 28.109

So you did come up with the correct answer, but maybe not for the right reasons, based on assumption. I did not carefully read each post in the thread, so you may have come up with the correct answer using the correct formula or logic.
 
Last edited:
"(because we know the angle of the arc given the right triangle)"

are you talking about the right triangle formed by the tank side walls? i thought that was irrelevant since the arc could be (and is) less then 90 degrees:

167670math_rofl.gif


oh and the reason you got 28.109 and I got 28.111 is that you used 37.123 to represent 37 1/8, whereas I used the actual decimal value (37.125).

if you know the radius then you know the diameter.

I didnt check, but I bet you could convert the equation I used ( c = 2(sqrt[h(2R - h)]) ) into the one you used.
 
Yes, the right angle formed by the tank walls gives us the chord length by simply solving for the length of the hypotenuse.

I know where the 28.109 came from :) and also know that the radius is half of the diameter (I am not sure what would make you think I didn't).

The equation you used may very well factor to the same thing i used. There are a dozen different ways to do it :) Like I said, I just skimmed through the thread and was not sure what method people used to derive the numbers they were coming up with. I don't enjoy reading math and trying to figure out what other people did, it brings back bad memories of word problems on tests. I just happen to know a few differnet ways to solve the problem and figured I would offer a few up so that others could learn something.
 
"Ryan:
The problem is that you don't know the diameter."

thats what made me think it lol.

"Yes, the right angle formed by the tank walls gives us the chord length by simply solving for the length of the hypotenuse. "

sure I get that... but we already knew the hypotenuse from the diagram drawn by the OP
 
Well, here's the realization of your hard work:

main.php

main.php

main.php


It worked! The radius is almost perfect. It seems slightly tighter on the sides than in the front, but that could have been my error. It's more than good enough for me.
Thanks again everyone, you rock!
 
Last edited:
WOW you did a really nice job putting that together.

a bedside tank too, very nice :) is it pretty quiet?

what plant is that in the foreground? it is SW right?


just a quick tip... i was watching new yankee workshop the other day and noticed that when they make a curved template like that, they will cut some 45 degree grooves into the back to give a better surface to clamp onto. but looks like it worked just fine how you did it regardless.
 
<a href=showthread.php?s=&postid=12060232#post12060232 target=_blank>Originally posted</a> by RyanBrucks
WOW you did a really nice job putting that together.

a bedside tank too, very nice :) is it pretty quiet?

what plant is that in the foreground? it is SW right?
Thanks!
Lol yeah, it's right next to my bed. I don't have much choice, seeing as I'm in an apartment with a roommate, and I don't trust my tanks with others around (especially with puffers) and we have people over all the time. I had to stuff 3 tanks into my little room.
I have the 20g reef-to-be on the other side of the bed (also just built this canopy):
main.php

Both tanks are really quiet, the bowfront is basically silent. The Eheim canister is great. I want to build a sump for th2 20g, but I'm afraid it will be too noisy.

The bowfront tank is actually brackish, ~ 1.010. Plants are just quality fakes :( It's tough to plant a brackish tank.
 
Back
Top