Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Problem solving

  1. Jul 11, 2004 #1
    Do you know some websites that I can learn to solve physics problem ?
    I know the concepts well but when I start to solve the problems, mess up :cry:
    I searched for this in google but I found just some suggestions on problem solving and not a complete and clear way.
     
    Last edited: Jul 11, 2004
  2. jcsd
  3. Jul 12, 2004 #2
    I believe this is the site for which you are searching

    You should post a specific description of your problem, that someone might help you
     
  4. Jul 12, 2004 #3
    I have no specific problem that I can't solve !!
    I generally when face a physics problem don't really know what to do :)
    Because I'm reading physics on my own and there is no instructor I am looking for a website to learn from , how to solve physics problem :smile: .
     
    Last edited: Jul 12, 2004
  5. Jul 12, 2004 #4
    I'm kinda in the same boat. I am trying to hack my way through Quantum Mechanics. I could really use an 'Idiots' guide.
     
  6. Jul 12, 2004 #5

    Gza

    User Avatar

    I feel your pain brother. :cry: I used to buy every book on problem solving I could get my hands on, but never really ended up any better than I was before. The sad truth, the truth your teacher probably tells you again and again when you inform him/her that you understand the concepts but can't solve the problems; is that you don't truly understand the concepts. Problems are probably(more like actually) the best way to gauge your understanding. My advice is to get your hands on tons of problems with solutions, and work your tail off(no peeking at the solutions until you finished the problem, or are ready to jump off a ten story building; even then...). I know it sucks, but the payoff is worth it.


    Here are my favorites; most of which are actually kind of fun in a twisted masochistic kind of way :smile: (they range from beginner to advanced, make sure to thouroughly study the solutions after working the problems to learn some good problem solving techniques)

    http://liquids.deas.harvard.edu/oleg/competition/

    http://electron6.phys.utk.edu/phys594/archives/marchives.htm

    http://star.tau.ac.il/QUIZ/
     
  7. Jul 12, 2004 #6
    My old physics teacher always recommended a process along the lines of the following:

    Consider a simple problem like "A 5kg wheeled trolley has a force of 10N applied to it. From a stationary position, how long does it take for the trolley to reach a speed of 10 m/s?

    1. Read through the problem once.
    2. Write out all the data values you have in a summary table. So, for the above, this would look like:

    m = 5kg​
    F = 10N​

    3. Work out what the unknown is that you have to find. In simple problems, this is obvious as the question will say "Find the acceleration." Otherwise, look at what they are asking for and work out what value you can find with the data you have, that the required output can be calculated from.

    So in the above example, to find how long it takes to reach a given speed, we need to divide that speed by the acceleration of the trolley, to find the number of seconds required. So our unknown element, for the moment, is the acceleration a.

    So we add to our data table
    a = x​
    to show the unknown.

    3. We then draw a simple diagram showing the trolley, labelled with m = 5kg, an arrow pointing forward to show the force F = 10N that is applied to the trolley and its direction, and an arrow pointing forward under the trolley to show the direction of its motion and label it with the acceleration a = x.

    These simple little diagrams are often the most helpful part.

    4. With the diagram and data table to present the problem in the most stripped down way, uncluttered with misleading words, it's usually easy to see what equation you need to connect them. (It is worth putting some time and effort into learning the equations by rote, so that they spring easily to mind.) Write it down without substituting any of the data yet; in this case:

    F = m.a​

    5. Rearrange the equation to give an output that will be the value you need:

    F/m = a​

    6. Substitute in the values from your data table and calculate the result for the unknown:

    10N/5kg = 2 m/(s^2)​

    7. Then, if necessary, use the output value to do the arithmetic to get to the final result, in this case, dividing the speed of interest (10 m/s) by the acceleration (2 m/(s^2)) to get the time taken:

    10 m/s / 2 m/(s^2) = 5 s​

    I hope that's some help.
     
  8. Jul 12, 2004 #7

    berkeman

    User Avatar

    Staff: Mentor

  9. Jul 13, 2004 #8
    Actually I think its modelling the equations from the language thats the hard part most of the time. There are cases when nothing strikes you though but they are (usually) rare. A large number of problems in general physics require you to set up correct equations from facts about physical behaviour (similar to Mathematical Modeling). Once the equation is set up, it is normally a breeze to solve it.

    But there is no hard and fast rule to set up the correct equations and reach correct conclusions. The more problems you do in peace, the smaller probability of encountering newer ones in war (er..I mean in exams).

    I would be interested to know if logical tactics to attack problems are known. I never had access to any really out of the world ones when I was at school and for a long time, I didn't know how to even start considering a differential element and making its freebody diagram!
     
  10. Jul 13, 2004 #9
    It is wroth mentioning here that the reason I started to search for a strait way for problem solving is the following problem :

    Suppose you fire a rifle bullet (1600 m/s) in a shooting gallery and hear the gong on the target ring 0.731 s later.
    Taking the speed of sound to be 330 m/s and assuming the bullet travels straight downrange at a constant speed, how far away is the target ?


    I spent about 3 hours on the problem without any success and I was getting confused more and more. But a guy in this forum solved it just in 5 lines :

    (1) L = v_{bullet}t_1
    And then for t_2, we have:
    (2) L = v_{sound}t_2
    But we know t_T = t_1 + t_2, so can you find an equation for t_T in terms of the speeds and distance?

    Then I thought something must be wrong with the way I solve problems.
    What do you think about the problem above ? It is so easy, isn't it ?
    I got angry for not solving such an easy problem.
    However, now I'm solving problems as everybody advised me here, at an speed or maybe velocity :) of 3 problems per day.
    Thank you again.
     
  11. Jul 14, 2004 #10

    Gza

    User Avatar

    I know i'm starting to sound like a bag of cliches between my last post in this one, but here goes, it's all a matter of practice. I remember while learning kinematics it seemed like the hardest damned thing I had ever done. I look back on the problems I struggled with then and laugh at how easy the problems look now. Don't let the fact you can't do a specific problem weigh you down. Move on to the next problems you can do, until you work up an adequate skill set enabling you to go back and do the ones you couldn't do before. And the funny thing about kinematics is that the whole thing will seem like a crazy mess until you formally study vector calculus (for me at least). So just rough it out for now, there's a light at the end of the tunnel.
     
  12. Jul 14, 2004 #11
    Okay, here is a way to solve many introductory physics problems involving forces and motion.

    1. Find the acceleration of the body in question. This is the key.

    There are only two ways of finding the acceleration.

    a. Apply Newton's second law by vector summing the forces and dividing by the mass of the body.

    or,

    b. Applying the equations of motion.

    2. Once you have solved for the acceleration, then solve the problem for the unknown.

    a. If you used Newton's second law to find the acceleration, then apply the equations of motion to find the unknown.

    or,

    b. If you used the equations of motion to find the acceleration, then apply Newton's second law to find the unknown.


    Above all, find the acceleration. As Sir Isacc Newton said, if you ain't got the acceleration, you ain't got ****.
     
  13. Jul 14, 2004 #12
    When you were at school there wasn't any PhysicsForums.com ? :biggrin:
    I found it today :
    ftp://ftp.ncsu.edu/pub/ncsu/beichner/RB/GOALPaper.pdf
    Thanks
     
    Last edited: Jul 14, 2004
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Problem solving
  1. Need a problem solved (Replies: 4)

  2. Help solving problem (Replies: 7)

Loading...