Terminal velocity

jaydeehess

Penultimate Amazing
Joined
Nov 15, 2006
Messages
20,849
Location
40 miles north of the border
I see that BAUT is going to be closing down 9/11 threads and concentrating on astronomy and space instead.

There is one thread there that concerns terminal velocity of a piano dropped from the top of the WTC. Prior to 9/11 this would not have anything to do with conspiracies but the thread may disappear. ,,, and so I bring you all the math question
For the terminal velocity of a piano I get;

Fg=ma
Fd=0.5 r A Cd v2

Fg=Fd at terminal velocity

For a piano (grand piano style) use;

A=2 m2m=200 Kg
Cd=1.14(that of a prism/triangle)
r(rho) for a lower part of the atmosphere = 1.2 Kg/m3a=9.8m2
So ;

200(9.8)=0.5(1.2)(2)(1.14) v2v=sqrt(1433)
v=37.8 m/second = 136 kilometers per hour = 85 MPH

This assumes that the piano does not tumble but remains in a flat fall.

[/quote]

But then it gets interesting when one wants to find the time to reach terminal velocity

Nicolas said:
s = 0.5*g*t^2 => t = sqrt(2*s/g) that's without air resistance, zero initial velocity. (s = height of the tower)

Drag D = 0.5*rho*v^2*S*Cd

Air density rho can be taken from standard atmosphere, or from comparable weather conditions. S is the projected surface of the object. Cd can be estimated from this page:
http://www.grc.nasa.gov/WWW/K-12/airplane/shaped.html

Now you've got a force W = m*g acting downwards, a force D acting upwards, the object falling due to N = W-D. The acceleration a = N/m.

acceleration a(t) = g - (0.5*rho*S*Cd*v(t)^2)/m

Integrating find v(t), integrating once more finds s(t). With s known, you find t_fall.

a(t) = g - (0.5*rho*S*Cd*v(t)^2)/m

a(t) =9.8- (0.5*1.2*2*1.14)v(t)^2)/200
a(t) = 9.8- (0.007)v(t)^2
Self check, substituting a=0 at terminal velocity
0 = 9.8 - (0.007)v(t)^2
v= sqrt(1400) =37.4 m/s (given rounding errors I think this is close enough to the above calc of terminal velocity)

So;

dv/dt = 9.8 - (0.007)(ds/dt)2

after that I get lost as it has been 30 years since calculus classes.

This bums me out since I realize this is a first year problem but I cannot seem to understand how I integrate this to find s(t) and t_ as Nicolas lays out
 
But then it gets interesting when one wants to find the time to reach terminal velocity

I don't suspect this is actually the question you are asking...

Using idealized equations you never reach terminal velocity as you get asymptotically close but never actually reach it. Terminal velocity is the limit of the velocity equation over time to infinity.

You are more interested in the closed form (non-differential) equation of distance as a function of only time, right?
 
I don't suspect this is actually the question you are asking...

Using idealized equations you never reach terminal velocity as you get asymptotically close but never actually reach it. Terminal velocity is the limit of the velocity equation over time to infinity.

You are more interested in the closed form (non-differential) equation of distance as a function of only time, right?

uhh, yeah:D

With terminal V I am using the equation that has fg=fdrag. In this case it results in a velocity of 37.4 m/s so what i am looking for would be the time to reach a velocity that approachs 37.4 m/s
 
uhh, yeah:D

With terminal V I am using the equation that has fg=fdrag. In this case it results in a velocity of 37.4 m/s so what i am looking for would be the time to reach a velocity that approachs 37.4 m/s

Well every time is a time that is approaching 37.4 :)

You will start at 0 and approach 37.4 asymptotically, never actually reaching it. You will reach 10, 30, 35, and 37.39999999999999 at finite and calculable times.

Keep in mind this is using the idealized equations which neglect a bunch of important second order effects.

Let me put your mind at ease, though, the following equation..
dv/dt = 9.8 - (0.007)(ds/dt)2
..is not something you can solve after you 1 year of calc. It's a differential equation and solving it is relatively non-trivial without having taken (and remembered) your dif-eqs class.

Wikipedia has the equations solved into non-differential form:
http://en.wikipedia.org/wiki/Free-fall

d167571dcbd9ce3feec5ec36bafb81a2.png
(note: your equation fits this form)
995d03fe98778d09723bd3572a157dd8.png

eed4ffb67ce6f75cf56baa1536def886.png

478e7c5f221aeb7dd644951dfe6fa689.png


Now, are you more interested in the answer, or the actual mathematical derivation? If it's the second, how comfortable are you with dif-eq? I'd need to review it briefly to work it out but I imagine it'll drop right out after a Laplace transform.
 
Last edited:
I did a little more research, and no, your equation doesn't fit the above form, because the above form uses a linear drag model and usually drag is modeled quadratically (as in your example).

Wikipedia also offers the following approximation for (I think) second-order drag which is what you are using:

e6faadb6a100cf62a6b7fc6f21081b2a.png


http://en.wikipedia.org/wiki/Drag_%28physics%29

Hopefully one of these equations gets you where you need to be. Not really sure your level of mathematical sophistication nor the level required for your problem. Let me know if any of this is at all useful.
 
Last edited:
Wikipedia has the equations solved into non-differential form:
http://en.wikipedia.org/wiki/Free-fall

[qimg]http://upload.wikimedia.org/math/d/1/6/d167571dcbd9ce3feec5ec36bafb81a2.png[/qimg] (note: your equation fits this form)
[qimg]http://upload.wikimedia.org/math/9/9/5/995d03fe98778d09723bd3572a157dd8.png[/qimg]
[qimg]http://upload.wikimedia.org/math/e/e/d/eed4ffb67ce6f75cf56baa1536def886.png[/qimg]
[qimg]http://upload.wikimedia.org/math/4/7/8/478e7c5f221aeb7dd644951dfe6fa689.png[/qimg]
Those equations don't look right to me even if we are only considering viscous drag. The first equation shows acceleration being proportional to mg, not g.

Anyways, using only the 'v squared' term I get an equation for velocity that looks like:

[latex]$$v = v_0 - v_t \frac{1-e^{-kt}}{1+e^{-kt}}[/latex]

[latex]$$k = \frac{2g}{v_t}[/latex]

Where v sub t represents terminal velocity.

Walt

Edit: Looking at Anti-sophist's post, the equations are the same. I missed the tanh simplification, so I could rewrite my equation as ...

[latex]$$v = v_0 - v_t\tanh{\frac{g}{v_t}t}[/latex]

Doing a piece wise integration in excel gives about 13.7 seconds to fall 417m. Hitting 37m/s velocity at around 8.7 seconds.

Edited again to add, now that we have it tanh form we can integrate the velocity to get distance.

[latex]$$x = x_0 + v_0 t - \frac{v_t^2}{g} \ln{|\cosh{\frac{g}{v_t}t}|}[/latex]
 
Last edited:
I'd thought I'd post it before someone else did. Ignore the initial velocity term in my equations. It is in the wrong place. However, the equations do work when v-nought is zero.
 
:eye-poppi

My math sophistication ends with me dropping out of a physics program at university during second year, in favor of a college level electronics course. That was in 1977.

So, much of this looks familiar but :confused: :confused: ........

It is indeed the numbers that I was primarily interested in but I do like to try to understand how they are arrived at. That , apparently, won't happen unless I enroll again:D

Using Walt's equation for distance and time to vt
I get distance traveled in the 8.7 seconds to 'terminal velocity' as 228 meters, a little more than halfway down the building. (x0 and v0 both being zero.

Interesting, a much more dense item such as a solid steel column would likely still be acellerating significantly(since it always will be acellerating abeit less so the farther it goes) when it hit the ground especially if it did not fall from a height of 417 meters but came from lower down.

Thanks guys.
 
Last edited:
Let me put your mind at ease, though, the following equation..
..is not something you can solve after you 1 year of calc..
It would be difficult, but not impossible.

It's of the form dv/dt=a^2-v^2 (where a is a constant, not acceleration).

[latex]dv/dt=a^2-v^2

dv/(a^2-v^2)=dt

(1/(a-v)+1/(a+v))dv/2a=dt

(ln((a+v)/(a-v)))/2a=t

(ln((a+v)/(a-v)))=2at

(a+v)/(a-v)=e^{2at}

a+v=e^{2at}(a-v)

v+e^{2at}v=e^{2at}a-a

v=a(e^{2at}-1)/(e^{2at}+1)=a(e^{at}-e^{-at})/(e^{at}+e^{-at})

v=a*tanh(at)

ds/dt=a*tanh(at)[/latex]

s=|cosh(at)|

Then there are some constants of integration to deal with that I ignored. Takes some inspration, but there's nothing there that shouldn't be covered in basic calculus.
 
Last edited:

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom