Two smoothing operations will be roughly additive. This description taken from the wikipedia article on Gaussian blurring describes this:

Applying multiple, successive gaussian blurs to an image has the same effect as applying a single, larger gaussian blur, whose radius is the square root of the sum of the squares of the blur radii that were actually applied. For example, applying successive gaussian blurs with radii of 6 and 8 gives the same results as applying a single gaussian blur of radius 10, since \sqrt{6^2 + 8^2} = 10.

So two successive 8mm blurs would be the same as applying sqrt(64+64) = 11.3 mm gaussian blur.

3dFWHMx is a program that was intended to be used to estimate smoothness from the noise of a time series. 3dFWHMx does have a -2difMAD option that has been used in the past for estimating smoothness from structural images, but it's not clear this method is effective or useful. You can search our message board for previous posting regarding this option. I haven't had much luck with it though on MRI structural images. 3dLocalstat with "-stat fwhm" gives me something a little more useful and gives somewhat reasonable numbers. This is a short script to play around with (just an ad hoc attempt to get an estimate for smoothness). You may also want to look at 3dBlurtoFWHM.

#!/bin/tcsh
# est_smooth.csh
#
# estimate smoothness for a structural image (if such a thing is possible)
#
# usage:
# est_smooth.csh dset nbhd_size
#
# example:
#
# 3dmerge -1blur_fwhm 8.0 -prefix anat_ns_blur2 anat_ns_blur1+orig
# tcsh est_smooth.csh anat_ns_blur1+orig 5
#

set dset = $1
set nbhd_size = $2
rm -f temp_localblur+orig.*

# estimate smoothness measure locally
3dLocalstat -stat FWHM -mask $dset -prefix temp_localblur -nbhd "RECT($nbhd_size,$nbhd_size, $nbhd_size)" $dset

set medx = `3dBrickStat -median -positive temp_localblur+orig'[0]'`
set medy = `3dBrickStat -median -positive temp_localblur+orig'[1]'`
set medz = `3dBrickStat -median -positive temp_localblur+orig'[2]'`

echo "Median blur estimates in x,y,z are $medx[2], $medy[2], $medz[2]"