Delta Function Integration: Justified or Fudging?

BOAS
Messages
546
Reaction score
19
Hello,

I feel like I am fudging these integrals a bit and would like some concrete guidance about what's going on.

1. Homework Statement

Evaluate ##I = \int_{-1}^{1} dx \delta'(x)e^3{x} ##

Homework Equations

The Attempt at a Solution


[/B]
I use integration by parts as follows,
##u = e^{3x}##
##du = 3e^{3x} dx##

##dv = \delta'(x)##
##v = \delta(x)##

##I = uv|^{1}_{-1} - \int^{1}_{-1} v du##

Evaluating ##uv## gives zero because the delta function is zero at both limits.

##I = - \int^{1}_{-1} dx \delta(x) 3e^{3x}##

Clearly, the only non zero value of this function is at ##x = 0##, making me think the answer to my problem is ##I = -3##, but this is the step that I think I'm 'fudging'. I feel like I'm disregarding the integration sign...

Is this justified?

Thanks for any guidance you can give.
 
Physics news on Phys.org
That's justified because ##\delta(x)## is centered in ##x=0##, which is contained by the integration limits.
 
That looks correct. It sometimes helps to think of a finite spiked function instead of delta. Let's say it's 0 outside ##[-a, a]## for some small a, and has a value of ##1/2a## in that interval. The integral is then approx: ##2a \frac{1}{2a} e^0 = e^0##
 
PeroK said:
That looks correct. It sometimes helps to think of a finite spiked function instead of delta. Let's say it's 0 outside ##[-a, a]## for some small a, and has a value of ##1/2a## in that interval. The integral is then approx: ##2a \frac{1}{2a} e^0 = e^0##

I don't quite follow.

Where do you evaluate the limits?

I see that the spiked function is ##\frac{1}{2a}## inside this interval, and since (I assume) it's centered on ##0##, I see where ##e^0## comes from, but not the ##2a##.
 
BOAS said:
I don't quite follow.

Where do you evaluate the limits?

I see that the spiked function is ##\frac{1}{2a}## inside this interval, and since (I assume) it's centered on ##0##, I see where ##e^0## comes from, but not the ##2a##.

2a is the width of the interval. An integral is the area under a curve!
 
PeroK said:
2a is the width of the interval. An integral is the area under a curve!

Ah!

I'm with you now. So an integral of a function multiplied by the dirac delta function is like evaluating the function itself at the point where the delta function is non-zero?
 
BOAS said:
Ah!

I'm with you now. So an integral of a function multiplied by the dirac delta function is like evaluating the function itself at the point where the delta function is non-zero?

That's exactly what it is. You can think of the delta function as the limit of a sequence of rectangular functions with area 1. That gives some justification for things.
 
PeroK said:
That's exactly what it is. You can think of the delta function as the limit of a sequence of rectangular functions with area 1. That gives some justification for things.

Cool - Thanks for your help!
 
Back
Top