The Student Room Group

Annoying vector question.

How do you find the vecor equation of the line that runs through the shortest distance between two non-collinear vectors.
I can find out the gradient using cross product but I dont know how to get a coordinate.

Apparently there are more than one method of doing these styles of questions and I cant understand any.

Can someone explain to me how you approach this.

Thank you :smile:
Reply 1
Try and use similtaneous eqns for the two collinear eqns? And the apply the r=b(lamda)+(1-lamda)a formula?
Reply 2
Original post by Khodu
Try and use similtaneous eqns for the two collinear eqns? And the apply the r=b(lamda)+(1-lamda)a formula?


how would that work?.

Also the two non-collinear lines cannot intersect.
Reply 4
Original post by imzir
doesnt have it :frown:


Do you have a question you're working on right now? post it?
Reply 6
Original post by imzir
how would that work?.

Also the two non-collinear lines cannot intersect.


sorry, just realised what I said.

Collinear means when they're both on the same line. So they both have the same direction vector?

http://www.maths.qmul.ac.uk/~sb/MTH4103/Chap06.pdf

6.2 has notes on collinear, hope that helps. I'm currently revising something else br otherwise I'd help. :yy: goodluck
Reply 7
l1 is r = 5i + j + 5k + s(i - j - 2k)
l2 is r = i + 11j + 2k + t(-4i - 14j + 2k )

part 1 ) find shortest distance - done

part 2 ) find vector equation of the line that contains the shortest distance found in part 1 - cant do.

Original post by boromir9111
..
(edited 12 years ago)
Original post by imzir
l1 is r = 5i + j + 5k + s(i - j - 2k)
l2 is r = i + 11j + 2k + t(-4i - 14j + 2k )

part 1 ) find shortest distance - done

part 2 ) find vector equation of the line that contains the shortest distance found in part 1 - cant do.


did you get 90/2sqr(259) for shorted distance?

Also wats the answer to part 2?
Reply 9
Original post by Freakonomics123
did you get 90/2sqr(259) for shorted distance?

Also wats the answer to part 2?


sorry - the answer isnt in the book.

No I didnt get that distance. Can you explain to me how you got that - i think im wrong.
Original post by imzir
l1 is r = 5i + j + 5k + s(i - j - 2k)
l2 is r = i + 11j + 2k + t(-4i - 14j + 2k )

part 1 ) find shortest distance - done

part 2 ) find vector equation of the line that contains the shortest distance found in part 1 - cant do.


http://www.netcomuk.co.uk/~jenolive/skew.html this may help you!
Reply 11
Original post by imzir
sorry - the answer isnt in the book.

No I didnt get that distance. Can you explain to me how you got that - i think im wrong.


The shortest distance between two skew lines can be calculated by this:

(ac)(b×d)b×d \dfrac{(\mathbf{a} - \mathbf{c}) \cdot (\mathbf{b} \times \mathbf{d})}{|\mathbf{b} \times \mathbf{d}|}, where a and c are the position vectors of the 2 lines and b and d are their direction vectors.
(edited 12 years ago)
Reply 12
Original post by dknt
The shortest distance between two skew lines can be calculated by this:

(ac)(b×d)b×d\dfrac{(\mathbf{a} - \mathbf{c}) \cdot (\mathbf{b} \times \mathbf{d})}{|\mathbf{b} \times \mathbf{d}|}, where a and c are the position vectors of the 2 lines and b and d are their direction vectors.


thats what i used. But what about the second part?
Reply 13


thats not the problem. The problem is the second part - actually finding the vector equation.
Original post by imzir
thats not the problem. The problem is the second part - actually finding the vector equation.


If your two lines are a+sb and c+td.

A direction vector of the line of shortest length is bxd

So, one way is, by going along the first line, and then across the shortest connecting line, you must get to the second line, hence:

a+sb + u(bxd) = c + td where u is a constant.

Solve for s and t, and hence get the two points on the shortest line, and derive its equation from them.

I expect there is a slicker method, but this is doing it from scratch.
Reply 15
Original post by ghostwalker
If your two lines are a+sb and c+td.

A direction vector of the line of shortest length is bxd

So, one way is, by going along the first line, and then across the shortest connecting line, you must get to the second line, hence:

a+sb + u(bxd) = c + td where u is a constant.

Solve for s and t, and hence get the two points on the shortest line, and derive its equation from them.

I expect there is a slicker method, but this is doing it from scratch.


Thanks that makes a lot of sense too (when I draw a diagram).

:biggrin:
Reply 16
Original post by ghostwalker
If your two lines are a+sb and c+td.

A direction vector of the line of shortest length is bxd

So, one way is, by going along the first line, and then across the shortest connecting line, you must get to the second line, hence:

a+sb + u(bxd) = c + td where u is a constant.

Solve for s and t, and hence get the two points on the shortest line, and derive its equation from them.

I expect there is a slicker method, but this is doing it from scratch.


Sorry do you mind explaining a bit more? How would I get s and t from there? Would we have to find u? :rolleyes:
Original post by dknt
Sorry do you mind explaining a bit more? How would I get s and t from there? Would we have to find u? :rolleyes:


By equating coefficients of i, then j, and then k, you end up with three equations in three unknowns.
You're not actually interested in u (since you've already worked out the shortest length), but if you wanted you can use u|bxd| as the shortest length.

Quick Reply

Latest