Method of Characteristics: Characteristic Curve Question

The name of the pictureThe name of the pictureThe name of the pictureClash Royale CLAN TAG#URR8PPP











up vote
0
down vote

favorite
1












I was browsing through some questions on Method of Characteristics and came across a couple solutions involving characteristic curves:
1. Solve Transport Equation using Method of Characteristics
2. Solve the PDE by method of characteristics.



This method involving the "system of characteristic differential equations" is different to the approach I was taught, and it makes a lot more sense.



The method we were told involves writing out a differential for each segment of the PDE, solving it, and applying boundary and/or initial conditions when necessary. It becomes messy and there are terms popping up, I don't really understand it, despite attempting to go through numerous example questions. I decided to study the 2 questions above.



My problem is with the second one, I just cannot understand it.




First family of characteristic curves, from $quad fracdtt^2=fracdxx^2 quadtoquad frac1t-frac1x=c_1$




My workout led to a different solution:




$$dx=fracx^2t^2dt$$
$$x=intfracx^2t^2dt$$




This allows me to divide both sides by $x^2$. Integrating $t^-2$ gives $-t^-1+C_1$. My answer is:




$$frac1x+frac1t=C_1$$




Then comes the second bit:




Second family of characteristic curves, from $quad fracdtt^2=fracdxx^2= fracfrac1tdt+frac1xdxfrac1tt^2+frac1xx^2 = fracfrac1tdt+frac1xdxt+x=fracdu(t+x)u$
$frac1tdt+frac1xdx=fracduu quadtoquad fracutx=c_2$




Which I do not understand at all.



I decided to do some algebra, and it seems that the pile comes from this:
$$fracdxx+fracdtt=0$$
Then dividing both sides by $(x+t)$, combining the two fractions, and subsequently equating it with $fracdu(t+x)u$.



I then tried to solve the equation:




$$fractdx+xdttx(x+t)=fracdu(t+x)u$$




Which gave me some natural logarithmic.
I multiplied both sides by $t+x$, and then integrated both sides:




$$fracdxx+fracdtt=fracduu$$
$$log(x)+log(t)-log(u)=C_2$$
Edit: I think this is wrong. I cannot integrate the equation with regards to 3 different differentials, right?




Raising this mess as the power of $e$ gets rid of the logs.
And, I am stuck. This looks completely different to the answer, and clearly wrong.



The first issue was likely a typo, since the negative sign was not carried through the calculations.



Any help regarding to the question is greatly appreciated. I tried looking online for material regarding to this approach and did not find anything useful. All the materials I can find does not involve "system of characteristic differential equations" but rather goes straight to various differentials. I think the second step has something to do with separation of variables.



Thanks!







share|cite|improve this question




















  • I found this article: www1.maths.leeds.ac.uk/~kersale/Teach/M3414/Notes/m3414_1.pdf I'll give it a read.
    – Hypergeometry
    Aug 28 at 14:26















up vote
0
down vote

favorite
1












I was browsing through some questions on Method of Characteristics and came across a couple solutions involving characteristic curves:
1. Solve Transport Equation using Method of Characteristics
2. Solve the PDE by method of characteristics.



This method involving the "system of characteristic differential equations" is different to the approach I was taught, and it makes a lot more sense.



The method we were told involves writing out a differential for each segment of the PDE, solving it, and applying boundary and/or initial conditions when necessary. It becomes messy and there are terms popping up, I don't really understand it, despite attempting to go through numerous example questions. I decided to study the 2 questions above.



My problem is with the second one, I just cannot understand it.




First family of characteristic curves, from $quad fracdtt^2=fracdxx^2 quadtoquad frac1t-frac1x=c_1$




My workout led to a different solution:




$$dx=fracx^2t^2dt$$
$$x=intfracx^2t^2dt$$




This allows me to divide both sides by $x^2$. Integrating $t^-2$ gives $-t^-1+C_1$. My answer is:




$$frac1x+frac1t=C_1$$




Then comes the second bit:




Second family of characteristic curves, from $quad fracdtt^2=fracdxx^2= fracfrac1tdt+frac1xdxfrac1tt^2+frac1xx^2 = fracfrac1tdt+frac1xdxt+x=fracdu(t+x)u$
$frac1tdt+frac1xdx=fracduu quadtoquad fracutx=c_2$




Which I do not understand at all.



I decided to do some algebra, and it seems that the pile comes from this:
$$fracdxx+fracdtt=0$$
Then dividing both sides by $(x+t)$, combining the two fractions, and subsequently equating it with $fracdu(t+x)u$.



I then tried to solve the equation:




$$fractdx+xdttx(x+t)=fracdu(t+x)u$$




Which gave me some natural logarithmic.
I multiplied both sides by $t+x$, and then integrated both sides:




$$fracdxx+fracdtt=fracduu$$
$$log(x)+log(t)-log(u)=C_2$$
Edit: I think this is wrong. I cannot integrate the equation with regards to 3 different differentials, right?




Raising this mess as the power of $e$ gets rid of the logs.
And, I am stuck. This looks completely different to the answer, and clearly wrong.



The first issue was likely a typo, since the negative sign was not carried through the calculations.



Any help regarding to the question is greatly appreciated. I tried looking online for material regarding to this approach and did not find anything useful. All the materials I can find does not involve "system of characteristic differential equations" but rather goes straight to various differentials. I think the second step has something to do with separation of variables.



Thanks!







share|cite|improve this question




















  • I found this article: www1.maths.leeds.ac.uk/~kersale/Teach/M3414/Notes/m3414_1.pdf I'll give it a read.
    – Hypergeometry
    Aug 28 at 14:26













up vote
0
down vote

favorite
1









up vote
0
down vote

favorite
1






1





I was browsing through some questions on Method of Characteristics and came across a couple solutions involving characteristic curves:
1. Solve Transport Equation using Method of Characteristics
2. Solve the PDE by method of characteristics.



This method involving the "system of characteristic differential equations" is different to the approach I was taught, and it makes a lot more sense.



The method we were told involves writing out a differential for each segment of the PDE, solving it, and applying boundary and/or initial conditions when necessary. It becomes messy and there are terms popping up, I don't really understand it, despite attempting to go through numerous example questions. I decided to study the 2 questions above.



My problem is with the second one, I just cannot understand it.




First family of characteristic curves, from $quad fracdtt^2=fracdxx^2 quadtoquad frac1t-frac1x=c_1$




My workout led to a different solution:




$$dx=fracx^2t^2dt$$
$$x=intfracx^2t^2dt$$




This allows me to divide both sides by $x^2$. Integrating $t^-2$ gives $-t^-1+C_1$. My answer is:




$$frac1x+frac1t=C_1$$




Then comes the second bit:




Second family of characteristic curves, from $quad fracdtt^2=fracdxx^2= fracfrac1tdt+frac1xdxfrac1tt^2+frac1xx^2 = fracfrac1tdt+frac1xdxt+x=fracdu(t+x)u$
$frac1tdt+frac1xdx=fracduu quadtoquad fracutx=c_2$




Which I do not understand at all.



I decided to do some algebra, and it seems that the pile comes from this:
$$fracdxx+fracdtt=0$$
Then dividing both sides by $(x+t)$, combining the two fractions, and subsequently equating it with $fracdu(t+x)u$.



I then tried to solve the equation:




$$fractdx+xdttx(x+t)=fracdu(t+x)u$$




Which gave me some natural logarithmic.
I multiplied both sides by $t+x$, and then integrated both sides:




$$fracdxx+fracdtt=fracduu$$
$$log(x)+log(t)-log(u)=C_2$$
Edit: I think this is wrong. I cannot integrate the equation with regards to 3 different differentials, right?




Raising this mess as the power of $e$ gets rid of the logs.
And, I am stuck. This looks completely different to the answer, and clearly wrong.



The first issue was likely a typo, since the negative sign was not carried through the calculations.



Any help regarding to the question is greatly appreciated. I tried looking online for material regarding to this approach and did not find anything useful. All the materials I can find does not involve "system of characteristic differential equations" but rather goes straight to various differentials. I think the second step has something to do with separation of variables.



Thanks!







share|cite|improve this question












I was browsing through some questions on Method of Characteristics and came across a couple solutions involving characteristic curves:
1. Solve Transport Equation using Method of Characteristics
2. Solve the PDE by method of characteristics.



This method involving the "system of characteristic differential equations" is different to the approach I was taught, and it makes a lot more sense.



The method we were told involves writing out a differential for each segment of the PDE, solving it, and applying boundary and/or initial conditions when necessary. It becomes messy and there are terms popping up, I don't really understand it, despite attempting to go through numerous example questions. I decided to study the 2 questions above.



My problem is with the second one, I just cannot understand it.




First family of characteristic curves, from $quad fracdtt^2=fracdxx^2 quadtoquad frac1t-frac1x=c_1$




My workout led to a different solution:




$$dx=fracx^2t^2dt$$
$$x=intfracx^2t^2dt$$




This allows me to divide both sides by $x^2$. Integrating $t^-2$ gives $-t^-1+C_1$. My answer is:




$$frac1x+frac1t=C_1$$




Then comes the second bit:




Second family of characteristic curves, from $quad fracdtt^2=fracdxx^2= fracfrac1tdt+frac1xdxfrac1tt^2+frac1xx^2 = fracfrac1tdt+frac1xdxt+x=fracdu(t+x)u$
$frac1tdt+frac1xdx=fracduu quadtoquad fracutx=c_2$




Which I do not understand at all.



I decided to do some algebra, and it seems that the pile comes from this:
$$fracdxx+fracdtt=0$$
Then dividing both sides by $(x+t)$, combining the two fractions, and subsequently equating it with $fracdu(t+x)u$.



I then tried to solve the equation:




$$fractdx+xdttx(x+t)=fracdu(t+x)u$$




Which gave me some natural logarithmic.
I multiplied both sides by $t+x$, and then integrated both sides:




$$fracdxx+fracdtt=fracduu$$
$$log(x)+log(t)-log(u)=C_2$$
Edit: I think this is wrong. I cannot integrate the equation with regards to 3 different differentials, right?




Raising this mess as the power of $e$ gets rid of the logs.
And, I am stuck. This looks completely different to the answer, and clearly wrong.



The first issue was likely a typo, since the negative sign was not carried through the calculations.



Any help regarding to the question is greatly appreciated. I tried looking online for material regarding to this approach and did not find anything useful. All the materials I can find does not involve "system of characteristic differential equations" but rather goes straight to various differentials. I think the second step has something to do with separation of variables.



Thanks!









share|cite|improve this question











share|cite|improve this question




share|cite|improve this question










asked Aug 28 at 14:11









Hypergeometry

607




607











  • I found this article: www1.maths.leeds.ac.uk/~kersale/Teach/M3414/Notes/m3414_1.pdf I'll give it a read.
    – Hypergeometry
    Aug 28 at 14:26

















  • I found this article: www1.maths.leeds.ac.uk/~kersale/Teach/M3414/Notes/m3414_1.pdf I'll give it a read.
    – Hypergeometry
    Aug 28 at 14:26
















I found this article: www1.maths.leeds.ac.uk/~kersale/Teach/M3414/Notes/m3414_1.pdf I'll give it a read.
– Hypergeometry
Aug 28 at 14:26





I found this article: www1.maths.leeds.ac.uk/~kersale/Teach/M3414/Notes/m3414_1.pdf I'll give it a read.
– Hypergeometry
Aug 28 at 14:26











1 Answer
1






active

oldest

votes

















up vote
0
down vote



accepted










First family: “This allows me to divide both sides by $x^2$.” No, you can't. The variables $x$ and $t$ are not independent, so you can't take $x^2$ outside of the $t$ integral.



What you have is
$$dt/t^2= dx/x^2 iff d(-1/t)=d(-1/x) iff d(1/t-1/x)=0 iff 1/t-1/x=c_1.$$



Second family: If $A=B$, then $(tA+xB)/(t+x)=(tA+xA)/(t+x)=A(t+x)/(t+x)=A$ (which also equals $B$). That is,
$$
A=B
quadimpliesquad
A = B = fractA+xBt+x
.
$$
The computation starts out by applying this identity with $A=dt/t^2$ and $B=dx/x^2$.



And then you get
$$
dt/t + dx/x = du/u
iff
d ln|t| + d ln|x| = d ln|u|
iff
d ln|tfracuxt|=0
iff
tfracuxt=c_2
.
$$
(If the logarithm of the absolute value of something is constant, then that something itself is constant too.)






share|cite|improve this answer




















  • Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
    – Hypergeometry
    Aug 28 at 14:49










  • Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
    – Hypergeometry
    Aug 28 at 15:07











  • You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
    – Hans Lundmark
    Aug 28 at 17:03










Your Answer




StackExchange.ifUsing("editor", function ()
return StackExchange.using("mathjaxEditing", function ()
StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix)
StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
);
);
, "mathjax-editing");

StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "69"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);

else
createEditor();

);

function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
convertImagesToLinks: true,
noModals: false,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
noCode: true, onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);



);













 

draft saved


draft discarded


















StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2897305%2fmethod-of-characteristics-characteristic-curve-question%23new-answer', 'question_page');

);

Post as a guest






























1 Answer
1






active

oldest

votes








1 Answer
1






active

oldest

votes









active

oldest

votes






active

oldest

votes








up vote
0
down vote



accepted










First family: “This allows me to divide both sides by $x^2$.” No, you can't. The variables $x$ and $t$ are not independent, so you can't take $x^2$ outside of the $t$ integral.



What you have is
$$dt/t^2= dx/x^2 iff d(-1/t)=d(-1/x) iff d(1/t-1/x)=0 iff 1/t-1/x=c_1.$$



Second family: If $A=B$, then $(tA+xB)/(t+x)=(tA+xA)/(t+x)=A(t+x)/(t+x)=A$ (which also equals $B$). That is,
$$
A=B
quadimpliesquad
A = B = fractA+xBt+x
.
$$
The computation starts out by applying this identity with $A=dt/t^2$ and $B=dx/x^2$.



And then you get
$$
dt/t + dx/x = du/u
iff
d ln|t| + d ln|x| = d ln|u|
iff
d ln|tfracuxt|=0
iff
tfracuxt=c_2
.
$$
(If the logarithm of the absolute value of something is constant, then that something itself is constant too.)






share|cite|improve this answer




















  • Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
    – Hypergeometry
    Aug 28 at 14:49










  • Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
    – Hypergeometry
    Aug 28 at 15:07











  • You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
    – Hans Lundmark
    Aug 28 at 17:03














up vote
0
down vote



accepted










First family: “This allows me to divide both sides by $x^2$.” No, you can't. The variables $x$ and $t$ are not independent, so you can't take $x^2$ outside of the $t$ integral.



What you have is
$$dt/t^2= dx/x^2 iff d(-1/t)=d(-1/x) iff d(1/t-1/x)=0 iff 1/t-1/x=c_1.$$



Second family: If $A=B$, then $(tA+xB)/(t+x)=(tA+xA)/(t+x)=A(t+x)/(t+x)=A$ (which also equals $B$). That is,
$$
A=B
quadimpliesquad
A = B = fractA+xBt+x
.
$$
The computation starts out by applying this identity with $A=dt/t^2$ and $B=dx/x^2$.



And then you get
$$
dt/t + dx/x = du/u
iff
d ln|t| + d ln|x| = d ln|u|
iff
d ln|tfracuxt|=0
iff
tfracuxt=c_2
.
$$
(If the logarithm of the absolute value of something is constant, then that something itself is constant too.)






share|cite|improve this answer




















  • Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
    – Hypergeometry
    Aug 28 at 14:49










  • Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
    – Hypergeometry
    Aug 28 at 15:07











  • You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
    – Hans Lundmark
    Aug 28 at 17:03












up vote
0
down vote



accepted







up vote
0
down vote



accepted






First family: “This allows me to divide both sides by $x^2$.” No, you can't. The variables $x$ and $t$ are not independent, so you can't take $x^2$ outside of the $t$ integral.



What you have is
$$dt/t^2= dx/x^2 iff d(-1/t)=d(-1/x) iff d(1/t-1/x)=0 iff 1/t-1/x=c_1.$$



Second family: If $A=B$, then $(tA+xB)/(t+x)=(tA+xA)/(t+x)=A(t+x)/(t+x)=A$ (which also equals $B$). That is,
$$
A=B
quadimpliesquad
A = B = fractA+xBt+x
.
$$
The computation starts out by applying this identity with $A=dt/t^2$ and $B=dx/x^2$.



And then you get
$$
dt/t + dx/x = du/u
iff
d ln|t| + d ln|x| = d ln|u|
iff
d ln|tfracuxt|=0
iff
tfracuxt=c_2
.
$$
(If the logarithm of the absolute value of something is constant, then that something itself is constant too.)






share|cite|improve this answer












First family: “This allows me to divide both sides by $x^2$.” No, you can't. The variables $x$ and $t$ are not independent, so you can't take $x^2$ outside of the $t$ integral.



What you have is
$$dt/t^2= dx/x^2 iff d(-1/t)=d(-1/x) iff d(1/t-1/x)=0 iff 1/t-1/x=c_1.$$



Second family: If $A=B$, then $(tA+xB)/(t+x)=(tA+xA)/(t+x)=A(t+x)/(t+x)=A$ (which also equals $B$). That is,
$$
A=B
quadimpliesquad
A = B = fractA+xBt+x
.
$$
The computation starts out by applying this identity with $A=dt/t^2$ and $B=dx/x^2$.



And then you get
$$
dt/t + dx/x = du/u
iff
d ln|t| + d ln|x| = d ln|u|
iff
d ln|tfracuxt|=0
iff
tfracuxt=c_2
.
$$
(If the logarithm of the absolute value of something is constant, then that something itself is constant too.)







share|cite|improve this answer












share|cite|improve this answer



share|cite|improve this answer










answered Aug 28 at 14:44









Hans Lundmark

33.5k564109




33.5k564109











  • Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
    – Hypergeometry
    Aug 28 at 14:49










  • Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
    – Hypergeometry
    Aug 28 at 15:07











  • You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
    – Hans Lundmark
    Aug 28 at 17:03
















  • Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
    – Hypergeometry
    Aug 28 at 14:49










  • Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
    – Hypergeometry
    Aug 28 at 15:07











  • You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
    – Hans Lundmark
    Aug 28 at 17:03















Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
– Hypergeometry
Aug 28 at 14:49




Thanks! I realised the division problem after attempting another problem, doing the same thing, and it led to nonsense. I'll try to comprehend the second family.
– Hypergeometry
Aug 28 at 14:49












Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
– Hypergeometry
Aug 28 at 15:07





Alright, the relation $A=B=fractA+xBt+x$ is essentially lowering the order two relationship to order one, right? Where does the term $t+x$ go though? There seems to be many clever tricks of this sort, our teacher simply told us to "remember them" is there a good paper online that I may go through to get a hold of it? Thanks for the help so far, greatly appreciated!
– Hypergeometry
Aug 28 at 15:07













You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
– Hans Lundmark
Aug 28 at 17:03




You multiply by $t+x$ on both sides. And yes, that $(tA+xB)/(t+x)$ thing was indeed clever; I don't know if I would have come up with that... You'll have to ask the original answerer where that came from!
– Hans Lundmark
Aug 28 at 17:03

















 

draft saved


draft discarded















































 


draft saved


draft discarded














StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2897305%2fmethod-of-characteristics-characteristic-curve-question%23new-answer', 'question_page');

);

Post as a guest













































































這個網誌中的熱門文章

How to combine Bézier curves to a surface?

Why am i infinitely getting the same tweet with the Twitter Search API?

Is there any way to eliminate the singular point to solve this integral by hand or by approximations?