Google Answers Logo
View Question
 
Q: Motion problem ( Answered 4 out of 5 stars,   1 Comment )
Question  
Subject: Motion problem
Category: Science > Physics
Asked by: meganerd-ga
List Price: $15.00
Posted: 20 Sep 2004 18:36 PDT
Expires: 20 Oct 2004 18:36 PDT
Question ID: 404018
An object at rest at point d0 is accelerated at rate a until the
velocity reaches velocity v1 at point d1. Velocity remains constant
until d2.

At point d2 the object is decelerated at rate -a to velocity v2 at
point d3. Velocity remains constant to d4

At position d4, the object is accelerated at acceleration a to d5,
then decelerated at rate -a so as to come to rest at desired position
d6 in the shortest possible time.

What is the time required to reach point d6? (as a formula)

To rephrase:

Acceleration and deceleration are at the same rate, but opposite direction.

Accelerate to velococity v1, coast to d2, decelerate to V2, coast to
d4, then come to rest at d6 as quickly as possible without exceeding
the accel/decel rate.


            v1            /\
         ________        /  \
        /        \      /    \
       /          \____/      \
      /                        \
     /              v2          \
    /                            \
   d0   d1      d2d3  d4  d5     d6
                                Time?
Answer  
Subject: Re: Motion problem
Answered By: livioflores-ga on 21 Sep 2004 08:39 PDT
Rated:4 out of 5 stars
 
Hi meganerd!!

For references (formulas, etc.) visit the following pages:
"Derivation of Kinematic Equations of Motion":
http://theory.uwinnipeg.ca/physics/onedim/node7.html#SECTION00251000000000000000

"Motion in one dimension":
http://physics.bu.edu/~duffy/py105/Motion1D.html

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

The graph:

Speed
           v1             /\
         ________        /  \
        /        \      /    \
       /          \____/      \
      /                        \
     /              v2          \
    /                            \
   d0   d1      d2d3  d4  d5     d6
                                    Time

is a velocity-time graph.

Call:
D1 = d1-d0 (distance between points d1 and d0)
D2 = d2-d1 (distance between points d2 and d1)
D3 = d3-d2 (distance between points d3 and d2)
D4 = d4-d3 (distance between points d4 and d3)
D5 = d5-d4 (distance between points d5 and d4)
D6 = d6-d5 (distance between points d6 and d5)


The distances Di are needed inputs like V1 and V2.

Time for D1: (t1) 

V1 = a*t1 

then:

t1 = V1/a

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

Time for D2: (t2)

D2 = V1*t2

Then

t2 = D2/V1

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

Time for D3: (t3)
 
V2-V1 = -a*t3

then

t3 = (V1-V2)/a

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

Time for D4: (t4)

D4 = V2*t4

Then

t4 = D4/V2

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

Time for D5: (t5)

Call V5 the final speed at d5, we have:

V5^2 = V2^2 + 2*a*D5

then 

V5 = sqrt(V2^2 + 2*a*D5)

V5-V2 = a*t5

then 

t5 = (V5-V2)/a = (sqrt(V2^2 + 2*a*D5) - V2) / a

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

Time for D6: (t6)

0 - V5 = -a*t6

then

t6 = V5/a = sqrt(V2^2 + 2*a*D5) / a

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

The total time T is:

T = t1 + t2 + t3 + t4 + t5 + t6 =
  = V1/a + D2/V1 + (V1-V2)/a + D4/V2 + t5 + t6 =

t5 + t6 = (V5 - V2)/a + V5/a  = (2*V5 - V2)/a =
        = (2*sqrt(V2^2 + 2*a*D5) - V2) / a

then

T = V1/a + D2/V1 + (V1-V2)/a + D4/V2 + (2*sqrt(V2^2 + 2*a*D5) - V2)/a =
  
  = [V1 + V1 - V2 + 2*sqrt(V2^2 + 2*a*D5) - V2]/a + D2/V1 + D4/V2 =
  
  = [2*V1 - 2*V2 + 2*sqrt(V2^2 + 2*a*D5)]/a + D2/V1 + D4/V2 =
 
  = 2*[V1 - V2 + sqrt(V2^2 + 2*a*D5)]/a + D2/V1 + D4/V2

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

I hope that this helps you. If you find that something is unclear,
need further assistance or if I misunderstood the problem don't
hesitate to request for an answer clarification; I will gladly respond
your requests.

Best regards.
livioflores-ga

Request for Answer Clarification by meganerd-ga on 21 Sep 2004 15:00 PDT
Knowns: d0 V1 d2 d4 d6
Unknowns: d1 d3 d5

On my first draft of this, only the known positions were labeled.
Later it was felt that labeling all the inflection points would
clarify things. No so.

I can solve this up to point d4 where it then gets a bit trickier.

At d4, I want to reach the goal at d6 ASAP. We know if I was starting
from zero velocity, I would want to accelerate for half the distance,
then decelerate to a stop at d6. I am however starting from some
positive velocity v1, so I don't know the optimum point (d5) to
decelerate.

Clarification of Answer by livioflores-ga on 21 Sep 2004 16:13 PDT
Hi again meganerd!!

Now the problem changes its face, doesn't it?
Don't worry, I will try to figure it by tonight.

Regards.
livioflores-ga

Request for Answer Clarification by meganerd-ga on 21 Sep 2004 16:55 PDT
This problem is not all hypothetical. Here's what I know in one
real-world situation:

a=0.175
v0=0 (starting velocity)
d0=0
v1=80
d2=145,000
v2=40
d4=170,000
d6=240,000
v6=0 (ending velocity)
units for acceleration are in encoder counts/sample
units for velocity are in encoder lines/sample
units for distance are in encoder lines
where a sample is 256 milliseconds
and a encoder line is 360/240,000 degrees

It might just as well be. . .
ft/sec/sec
ft/sec
ft
seconds

Request for Answer Clarification by meganerd-ga on 21 Sep 2004 16:58 PDT
Make that accelleration in encoder lines/sample/sample.

Clarification of Answer by livioflores-ga on 21 Sep 2004 23:58 PDT
Hi!!

hfshaw gave you an analitical solution, I will give you a geometric one.
Speed
                          /\
                         /  \
                        /    \
                   ____/_ _ _ \_
                              |\
                    V2        | \
                              |  \
                      d4  d5  d5' d6
                                   
                                        Time


"At d4, I want to reach the goal at d6 ASAP. We know if I was starting
from zero velocity, I would want to accelerate for half the distance,
then decelerate to a stop at d6. I am however starting from some
positive velocity v1, so I don't know the optimum point (d5) to
decelerate."

At this point I must agree with hfshaw, you have only one way to draw
the last part of the graph with a given slope (a) and given V2 and
final V = 0.

Knowing that, consider the point at the deceleration part when you
reach again the "initial" speed V2 and call it d5'.
From this point to the rest point d6 we have that:

t6 = V2/a 
   
   

And the distance covered is:

X5 = V2^2/2.a 

then

d5' = d6 - V2^2/2.a


Now you need to solve:

Speed
                          /\
                         /  \
                        /    \
                   ____/_ _ _ \_
                    V2                     
         
                      d4  d5  d5'
                                     Time


Note that we know d5', then the total distance for this part of the problem is:

D5' = d5' - d4  ;

its correspondent time is t5'

and

D5 = d5-d4 = D5'/2


Call V5 the final speed at d5, we have:

V5^2 = V2^2 + 2*a*D5

then 

V5 = sqrt(V2^2 + 2*a*D5)

V5-V2 = a*t5

then 

t5 = (V5-V2)/a = (sqrt(V2^2 + 2*a*D5) - V2) / a

By symmetry:

t5' = 2*t5 = 2*(sqrt(V2^2 + 2*a*D5) - V2) / a


Call T6 the total time for this part of the problem:

T6 = t5' + t6 =
   = 2*(sqrt(V2^2 + 2*a*D5) - V2) / a + V2/a



Using the data (and considering it in metrics units):

d5' = d6 - V2^2/2.a = 
    = 240,000 - 40^2/2*0.175 =
    = 240,000 - 9142.86 = 
    = 230857.14

D5' = d5' - d4 =
    = 230,857.14 - 170,000 =  
    = 60857.14

D5 = D5'/2 = 60857.14/2 = 30428.57 

V5 = sqrt(V2^2 + 2*a*D5) =
   = sqrt(1600 + 2*0.175*30428.57) =
   = 110.68

t5 = (V5-V2)/a =
   = (110.68 - 40)/0.175 =
   = 70.68/0.175 =
   = 403.88

t5' = 807.77


t6 = V2/a =
   = 40/0.175 =
   = 228.57

T6 = t5' + t6 =
   = 807.77 + 228.57 =
   = 1036.34

Clarification of Answer by livioflores-ga on 21 Sep 2004 23:59 PDT
I hope that the previous clarification helps you. Feel free to request
for a clarification if it is needed.


Best regards.
livioflores-ga

Request for Answer Clarification by meganerd-ga on 22 Sep 2004 14:13 PDT
Yes, I see there is only one solution that meets all the constraints.

I follow your logic for determining the time required to get from d4 to d6. 

There was a minor aritmetic error at one point. 40^2/2*0.175= 4571.43
lines, not 9142.86, but the logic is fine.

If I make this correction, the result correlates closely with the
real-world response of this system which is burdened with litle things
like overshoot and friction.

Could you represent the time to perform the whole move, from d0 to d6
as a single equation? My math skills are just as rusty as my physics
after 40 years of neglect.

Total Time = Equation?

Clarification of Answer by livioflores-ga on 22 Sep 2004 23:41 PDT
Hi again meganerd!!


Excuse me for the aritmetic error, I forgot to divide by 2.

"Could you represent the time to perform the whole move, from d0 to d6
as a single equation?"

I think that I can, basically is the same of the original problem with
some corrections:

We had:

T = t1 + t2 + t3 + t4 + t5 + t6 =
  = V1/a + D2/V1 + (V1-V2)/a + D4/V2 + t5 + t6 

The first part [V1/a + D2/V1 + (V1-V2)/a + D4/V2] is the time until
you reach d4 and the second part [t5 + t6] is the time from d4 to d6.


Let see the first part, we have:

[V1/a + D2/V1 + (V1-V2)/a + D4/V2]

where:
D2 = d2-->d1  (distance between d2 and d1)
D4 = d4-->d3  (distance between d4 and d3)

Recall the data:
Knowns: d0 V1 V2 d2 d4 d6
Unknowns: d1 d3 d5

Call:
D1 = d1-->d0  (distance between d1 and d0) unknown
D3 = d3-->d2  (distance between d3 and d2) unknown

D2' = d2-->d0 (distance between d2 and d0) known
D4' = d4-->d2 (distance between d4 and d2) known

Then:

D2 = D2' - D1
D4 = D4' - D3

We need to find D1 and D3:

D1 = d1-d0 = 1/2*V1^2*1/a = 
   = V1^2/2.a

D2 = D2' - (V1^2)/(2.a)


D3 = d3-d2 = 1/2*(V2+V1)*t3 =
           = 1/2*(V2+V1)*(V1-V2)/a =
           = (V1^2 - V2^2)/(2.a) =
then:

D4 = D4' - (V1^2)/(2.a) + (V2^2)/(2.a)

Then:

D2/V1 = [D2' - (V1^2)/(2.a)]/V1 =
      = [D2'/V1 - V1/(2.a)]

and

D4/V2 = [D4' - (V1^2 - V2^2)/(2.a)]/V2 =
      = [D4'/V2 - (V1^2 - V2^2)/(2.a.V2)] =
      = [D4'/V2 + V2/(2.a) - (V1^2)/(2.a.V2)]


Now we can rewrite the first part of the equation [V1/a + D2/V1 +
(V1-V2)/a + D4/V2] as:

[V1/a + D2'/V1 - V1/(2.a) + (V1-V2)/a + D4'/V2 + V2/(2.a) - 
(V1^2)/(2.a.V2)] =

= [V1/a + D2'/V1 + V1/(2.a) + D4'/V2 - V2/(2.a) - (V1^2)/(2.a.V2)] = 

= [3/2*V1/a + D2'/V1 + D4'/V2 - V2/(2.a) - (V1^2)/(2.a.V2)] = T4


For the second part we have:

D6' = d6-->d4  (distance between d6 and d4) known

d5' is the point at the deceleration part in which you reach again the
"initial" speed V2.

d5 is the point in wich we stop the acceleration and start the final
deceleration (midpoint between d4 and d5').
D5 = d5-->d4 ((distance between d5 and d4) unknown

D5 = 1/2*(d5'-->d4) 

X5 = d6-->d5' = V2^2/(2.a)

d5'-->d4 = D6' - X5 = D6' - V2^2/(2.a)

then:

D5 = (d5'-->d4)/2 =
   = 1/2*(D6' - V2^2/2.a)


Calling T6 the total time for this second part of the problem we have:

T6 = 2*(sqrt(V2^2 + 2*a*D5) - V2)/a + V2/a =
   = 2*(sqrt(V2^2 + 2*a*[1/2*(D6' - V2^2/2.a)]) - V2)/a + V2/a =
   = 2*(sqrt(V2^2 + a*D6' - V2^2/2) - V2)/a + V2/a =
   = 2*(sqrt(V2^2/2 + a*D6') - V2)/a + V2/a =
   = 2*sqrt(V2^2/2 + a*D6')/a - 2*V2/a + V2/a =
   = 2*sqrt(V2^2/2 + a*D6')/a - V2/a =


Putting all together:

T = T4 + T6 =
  = 3/2*V1/a + D2'/V1 + D4'/V2 - V2/(2.a) - (V1^2)/(2.a.V2) +
    + 2*sqrt(V2^2/2 + a*D6')/a - V2/a =

T = 3/2*V1/a + D2'/V1 + D4'/V2 - 3/2* V2/a - (V1^2)/(2.a.V2) +
    + 2*sqrt(V2^2/2 + a*D6')/a =

T = [3/2*(V1-V2) + 2*sqrt(V2^2/2 + a*D6') - (V1^2)/(2.V2)]/a +
    +  D2'/V1 + D4'/V2   =
  


This is the equation that I found, note that may be there are
different ways to express the same equation depending on the way that
you associate terms and simplify it. I also hope that there are no
mistakes and/or typos on it, if that happens, please excuse me and
consider how large the equation is, it is easy to make a mistake.

Clarification of Answer by livioflores-ga on 22 Sep 2004 23:45 PDT
Thank you for the rating and the generous tip.

I hope that the equation that I found works fine, please let me know
if you find an error, typo or something unclear on it. I will gladly
give you further assistance on this if you need it.


Best regards.
livioflores-ga
meganerd-ga rated this answer:4 out of 5 stars and gave an additional tip of: $10.00
My statement of the problem was not as clear as it should have been.
I'm still hoping to see the final equation.

Comments  
Subject: Re: Motion problem
From: hfshaw-ga on 21 Sep 2004 21:40 PDT
 
As I understand it, you want to answer the question: "Given a starting
velocity (v2), how long (t_a) should I accelerate at a constant
acceleration (a), and how long (t_d) should I then decelerate at a
constant negative acceleration (-a), if I want to travel a specified
distance (D = d6 - d4) and have zero velocity when I have finished
traveling that distance?"

You have phrased the problem in terms of the distance, d5, but if we
know the t_a and the starting velocity, we can always calculate the
distance travelled.

In addition, you have phrased the problem as though it is an
optimization problem, in which you are able to choose how long to
accelerate in order to minimize the total time.  In fact, given an
initial velocity, a distance that must be traveled, and an
acceleration, there is only one (or zero) possible solutions to this
problem.  There are no solutions for cases in which the initial
velocity, v2 > (2*D*a)^1/2.  When v2<=(2*D*a)^1/2, there is exactly
one solution.

Let:
t_a be the time during which one accelerates
t_d be the time during which one decelerates
v2 be the initial velocity (at d4 in your diagram)
v5 be the velocity after accelerating for a time t_a (the velocity at
the point d5 in your diagram)
v6= 0 be the velocity at d6
D = d6 - d4 be the distance that must be covered.


We have that:
v5 = v2 + a*t_a
v6 = 0 = v5 - a*t_d 

Eliminating v5 between these two equations yields
0 = v2+a*t_a - a*t_d

Rearranging to solve for t_d yields:
t_d = (v2 + a*t_a)/a


We also know that:
D = v2*t_a + 1/2*a*(t_a)^2 + v5*t_d - 1/2*a*(t_d)^2

where the first two terms on the r.h.s. give the distance travelled
during the acceleration phase, and the last two terms give the
distance travelled during the deceleration phase.

Substituting the above expressions for v5 and t_d into the distance
equation yields (after a little algebra):
D = 2*v2*t_a + ((v2)^2)/(2*a) + a*(t_a)^2

Using the quadratic formula to solve for t_a yields:
t_a = {-v2 +/- 1/2*(2*(v2)^2 + 4*D*a)^1/2}/a   

because t_a must be positive, we can immediately eliminate the root
involving the negative sign.

So, in order to meet the constraints you have specified, one should accelerate for

t_a = {-v2 + 1/2*[2*(v2)^2 + 4*D*a]^1/2}/a

Note that only solutions for which t_a> 0 are physically reasonable
(i.e., we can't accelerate for a negative amount of time once we're at
point d4).  That means that:

0<=  {-v2 + 1/2*[2*(v2)^2 + 4*D*a]^1/2}/a

Solving this for v2 yields the constraint v2 <= (2*D*a)^1/2 that I
mentioned at the beginning of this comment.

For your case, in which (s stands for sample):
a=0.175 lines/s^2
v2=40 lines/s
d4=170,000 lines
d6=240,000 lines

We have that:

D = 70,000 lines
t_a = 424.209 s
d5 = 32714.3 lines
t_d = 652.781 s

Important Disclaimer: Answers and comments provided on Google Answers are general information, and are not intended to substitute for informed professional medical, psychiatric, psychological, tax, legal, investment, accounting, or other professional advice. Google does not endorse, and expressly disclaims liability for any product, manufacturer, distributor, service or service provider mentioned or any opinion expressed in answers or comments. Please read carefully the Google Answers Terms of Service.

If you feel that you have found inappropriate content, please let us know by emailing us at answers-support@google.com with the question ID listed above. Thank you.
Search Google Answers for
Google Answers  


Google Home - Answers FAQ - Terms of Service - Privacy Policy