Announcement

Collapse
No announcement yet.

F.B Johnson's Paper

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • F.B Johnson's Paper

    I've been studying the attached paper - "A Pulsed Bomb Locator" by F.B. Johnson M.A.

    Equation 1 (at the bottom of page 2):



    This represents the magnetic field at a buried spherical target.

    In MKS units (from John Alldred's comments):



    OK - this is all very well, but what is the origin of this equation?

    From the Biot-Savart Law:



    i.e. The field of dB due to a small element dl of the circle, centered at A (a point on the circumference of the coil) is given by:



    where:
    x = distance along the coil axis
    r = radius of coil
    dl = infinitesimal length of conductor carrying current I
    = unit vector specifying distance to the field point
    = permeability of free space

    There are two components to the field. One is along the coil axis, and the other is perpendicular to the axis. Hence, the total magnetic field at a distance along the coil axis is:



    which now starts to bear some resemblance to equation 1.

    If there are T turns on the coil, then:



    and (since we have:



    which is John Alldred's MKS version of equation 1.


    So far, so good ... but we then move on to equation 2:



    where:
    I = the intensity of magnetization of a sphere (target object)
    = decay time of eddy currents in the target

    which is based on the assumption that the switch-off is less than one-tenth of the decay time of eddy currents in the target object.

    Despite repeated attempts, I am failing to derive this formula, although it appears that the volume of a sphere is in there somewhere.

    And then we have equation 3:



    where:
    a = radius of target sphere [cms]
    = conductivity of material [emu]

    John Alldred converted this to MKS units as follows:



    I'm not certain, but equation 3 seems somewhat suspect to me.

    For example ... let's take a small sphere of pure gold with a diameter of 5mm. In this case we have:

    a = 2.5 x
    = 4.10 x

    This yields (from the MKS version of equation 3) a of 32.6us ... which seems too high to me.

    If anyone fancies some homework, then see if you can derive equation 2, and figure out what's wrong with equation 3.

    I did think this might be useful to compare with the work that Green has been doing in determining the time constants of some standard metal targets, but now I'm confused.
    Attached Files
    Last edited by Qiaozhi; 11-20-2016, 09:56 PM.

  • #2
    What does MKS stand for ?

    Comment


    • #3
      Originally posted by greylourie View Post
      What does MKS stand for ?
      The MKS system of units is an international system of measurement used in engineering and physics.
      Length in meters, mass in kilograms, and time in seconds.
      It's also known as the International System of Units (SI units).

      The old system was known as CGS (centimeter, gram, second).

      Here's a comparison between the CGS and MKS systems ->
      https://www.unc.edu/~rowlett/units/cgsmks.html

      Comment


      • #4
        Metric/S.I is familiar. MKS and CGS were new terms to me. Thanks for clearing that up.

        Comment


        • #5
          Originally posted by Qiaozhi View Post
          For example ... let's take a small sphere of pure gold with a diameter of 5mm. In this case we have:

          a = 2.5 x
          = 4.10 x

          This yields (from the MKS version of equation 3) a of 32.6us ... which seems too high to me.
          For a given shape, the higher the conductivity the longer the time constant. I've been experimenting with 22kt gold beads and seen decays in this order of magnitude.

          Comment


          • #6
            Originally posted by Teleno View Post
            For a given shape, the higher the conductivity the longer the time constant. I've been experimenting with 22kt gold beads and seen decays in this order of magnitude.
            So, in your opinion, does this equation match reality?

            Perhaps the assumption of using pure gold is the problem, since nuggets contain other impurities which will affect the conductivity.

            If we do the same calculation for a 5mm diameter sphere of aluminium, where x , we have a of 27.85us, which seems more reasonable.
            Last edited by Qiaozhi; 03-15-2016, 10:46 PM.

            Comment


            • #7
              [which is based on the assumption that the switch-off is less than one-tenth of the decay time of eddy currents in the target object]

              What is switch off time?

              Comment


              • #8
                Originally posted by green View Post
                [which is based on the assumption that the switch-off is less than one-tenth of the decay time of eddy currents in the target object]

                What is switch off time?
                The switch-off time is not included in the equation. As [apparently] it "can be shown" that , so long as the time-constant of switch-off is less than about one-tenth of the decay time of eddy currents in the object, the behaviour of the eddy currents is practically the same as if the field had been removed instantaneously, and the exact shape of the switch-off is immaterial.

                You also have to remember that this paper was written in 1956, and the switching element was a thyratron. Also, it was only suitable for finding large buried targets with time constants in the milli-second range. Of course, this fact is evident in the title of the paper - "A Pulsed Bomb Locator". In the text, it states that the test object dimensions were not small in comparison to the coil dimensions, and the object of interest was a 100 lb bomb.

                In the conclusions, an example is given of a 200 cm radius coil (i.e. 4m diameter), with 600 turns, and a current of 25A, returned a signal of 10uV at 7.5ms after switch-off. This was from a 100 lb bomb casing that was coaxial with the coil, and at a depth of 45 feet.

                However, the equations should still be relevant for our own purposes, as long as the target object shares a common axis with the coil.

                Comment


                • #9
                  Originally posted by Qiaozhi View Post
                  The switch-off time is not included in the equation. As [apparently] it "can be shown" that , so long as the time-constant of switch-off is less than about one-tenth of the decay time of eddy currents in the object, the behaviour of the eddy currents is practically the same as if the field had been removed instantaneously, and the exact shape of the switch-off is immaterial.

                  You also have to remember that this paper was written in 1956, and the switching element was a thyratron. Also, it was only suitable for finding large buried targets with time constants in the milli-second range. Of course, this fact is evident in the title of the paper - "A Pulsed Bomb Locator". In the text, it states that the test object dimensions were not small in comparison to the coil dimensions, and the object of interest was a 100 lb bomb.

                  In the conclusions, an example is given of a 200 cm radius coil (i.e. 4m diameter), with 600 turns, and a current of 25A, returned a signal of 10uV at 7.5ms after switch-off. This was from a 100 lb bomb casing that was coaxial with the coil, and at a depth of 45 feet.

                  However, the equations should still be relevant for our own purposes, as long as the target object shares a common axis with the coil.
                  Qiaozhi and all interested,

                  The critical aspect of this topic is that the coil discharge time constant is based on the effective value of the damping resistor(Rd). This effective Rd value includes the input resistor in parallel with Rd (428 ohm Rd in parallel with 1000 ohms equals 300 ohms), typically about 1000 ohms for an effective Rd in the range of 300 ohms to 600 ohms down to the clamping diode voltage of about .6V then it is just the value of Rd. This effect puts a slight kink in the coil discharge curve for mono coil PI machines. Assume that the coil is 300 uH and the effective Rd is 300 ohms then the coil discharge TC is 1uS which would then stimulate a target of 10 uS very well but will there be enough signal left to detect at a delay of 10 uS? This is why the effective value of Rd, coil, coax and TX circuit capacitance is critical for detecting small, quick decaying targets. You want the coil discharge to be as close to vertical as possible for detecting the smallest low TC targets.

                  Knowing the TC of your primary sought after target allows you to do some reverse engineering to select the best coil discharge characteristics that match your desired target. Beach hunters looking for coins at 12 plus inches depth would do best with a 12 to 15 inch diameter mono coil but might pass over small gold targets like thin gold chains or small ear rings. Unbroken gold rings will come in like US nickels at a good depth due to their size and longer TC. Another problem facing beach hunters is the coil response to salt wet sand with delays below about 15 uS. Here, using smaller DD type search coils might allow lower delay settings with less response to wet sand and give more potential for finding smaller gold targets but at the expense of covering less ground with each coil swing.

                  Knowing a little theory about target TC, coil discharge TC and hunting environment can go a long way to optimizing equipment selection and hunting techniques.

                  i hope this helps?

                  Joseph J. Rogowski

                  Comment


                  • #10
                    Charted a small lead and aluminum ball. Zapped the Tx circuit yesterday, got some oscillation that shouldn't effect the decay curve after repairing circuit. The lead ball charts exponential, the aluminum doesn't until after it's TC time(maybe).
                    Attached Files

                    Comment


                    • #11
                      Originally posted by green View Post
                      Charted a small lead and aluminum ball. Zapped the Tx circuit yesterday, got some oscillation that shouldn't effect the decay curve after repairing circuit. The lead ball charts exponential, the aluminum doesn't until after it's TC time(maybe).
                      Using equation 3:



                      we have:

                      Target #1 (round lead split shot pinched closed) 5.3mm diameter. for lead = 4.55 x
                      = 4.1us versus 4.3us (measured)

                      Target #2 (round aluminium ball) 5.9mm diameter. for aluminium = 3.50 x
                      = 38.8us versus 25us (measured)

                      Regarding target #2, the measured result is probably being skewed by the initial [faster] decay shown in the curve.

                      It appears that Equation 3 is not​ as suspect as I first surmised.
                      Last edited by Qiaozhi; 10-05-2015, 10:43 PM. Reason: Slight error in calculation for target 2

                      Comment


                      • #12
                        By the way, if anyone of you want to understand the mysteries of LaTeX code, then here's a short lesson:

                        Firstly, you need to enclose the equation using the delimiters: [latex] - put equation here - [/latex]

                        Special characters: \sigma \tau \mu and \pi etc. ...

                        Fractions: \frac{x}{y} ...

                        Exponent: 10^6 10^{-7} \pi^2 ...

                        Subscript: H_o ...

                        Whitespace: \

                        Here's a couple of examples to make things clearer:

                        Example 1



                        LaTeX code is: [latex]V\ =\ \frac{4}{3}\ \pi\ r^3[/latex]

                        Example 2



                        LaTeX code is: [latex]e^{\frac{-t}{\tau}}[/latex]


                        Simple really.
                        Don't be shy ... try it!

                        Hint: Use the Advanced Edit Mode, and then use Preview Post to check for any errors before posting.

                        Comment


                        • #13
                          Originally posted by Qiaozhi View Post
                          Using equation 3:



                          we have:

                          Target #1 (round lead split shot pinched closed) 5.3mm diameter. for lead = 4.55 x
                          = 4.07us versus 4.3us (measured)

                          Target #2 (round aluminium ball) 5.9mm diameter. for aluminium = 3.50 x
                          = 39.78us versus 25us (measured)

                          Regarding target #2, the measured result is probably being skewed by the initial [faster] decay shown in the curve.

                          It appears that Equation 3 is not​ as suspect as I first surmised.
                          I'm having trouble with the simple math. Could you enter all the numbers in the formula as an example? I charted 3 more lead balls starting the decay curve much latter. Including the charts to show what I did. Won't be necessary for other time constants if they decay exponential and start time is included. Example: 30,4mm lead ball, TC=120usec starting at 200usec. Tried the 5.9mm ball again, TC= 37.3usec after 50usec, closer to what you calculated.
                          Attached Files

                          Comment


                          • #14
                            Originally posted by green View Post
                            I'm having trouble with the simple math. Could you enter all the numbers in the formula as an example? I charted 3 more lead balls starting the decay curve much latter. Including the charts to show what I did. Won't be necessary for other time constants if they decay exponential and start time is included. Example: 30,4mm lead ball, TC=120usec starting at 200usec. Tried the 5.9mm ball again, TC= 37.3usec after 50usec, closer to what you calculated.
                            The equation is:

                            Hence, for the examples supplied:



                            and:

                            1) 24.1mm diameter: = 84.1us ... measured as 87us

                            2) 30.4mm diameter: = 133.8us ... measured as 120us

                            3) 36.4mm diameter: = 191.9us ... measured as 190us


                            Pretty damn close!

                            Comment


                            • #15
                              Qiaozhi, Green
                              Great work!!!
                              Thank you,
                              Chet

                              Comment

                              Working...
                              X