Page 508 - High Power Laser Handbook

476 Fi b er L a s er s Pulsed Fiber Lasers 477

25
24
shrinkage and intermodal mixing, which should enable tight
bending with concomitant preservation of modal purity and low
NLE. However, the HOM beam quality (BQ) factor is modest (e.g.,
M for LP is greater than 6.7, according to Ref. 26), which, in high-
2
04
power applications, requires a lossy bulk mode adaptor to recover
good output BQ. In addition, the HOM peak irradiance is locally
much higher than LP (e.g., a 2100-μm -area LP has the same peak
2
07
01
irradiance as a 316-μm -area LP ), which poses a bulk-damage haz-
2
01
ard at high peak power. Finally, energy extraction efficiency and
multimode behavior issues may arise due to spatial hole burning.
As for core enlargement, the need for minimization of the fiber
length toward NLE containment poses important fiber design and
laser architecture challenges. In some cases, shortening the fiber leads
to optical efficiency degradation caused by incomplete pump absorp-
tion. This problem can be avoided in two ways: One is to increase the
rare-earth-dopant density. However, this technique is hardly an open-
ended solution due to the onset of doping-ion clustering at high
enough doping concentrations, resulting in excited-state quenching
and ensuing reduction of the excited-state lifetime. The second
approach is to increase the pump and doped-core overlap, which can
be done by design via the geometric increase of the fiber core:cladding
area ratio in double-clad fibers. This approach may result in reduced
brightness acceptance in the fiber pump cladding and has only recently
become a truly viable option due to the ever-increasing bright-
ness of pump diode lasers, which currently exceeds 10 MW/(cm sr)
2
within commercially available, fiber-delivered, angle- or polarization-
multiplexed single-stripe diodes.
Even assuming that effective pump absorption has been obtained
from maximization of such core:cladding area ratio, any net reduc-
tion of the doped area volume (i.e., reduction in the total number of
rare earth ions) ultimately results in a lower achievable small-signal
gain, simply due to the reduced energy storage capacity. Accordingly,
the extractable energy (proportional to the small-signal gain) will be
degraded, which means that when used as an amplifier, such fiber
must be seeded at high power for acceptably efficient operation.
27
Yet another issue with short-fiber amplifiers is thermal manage-
ment. Fibers are often referred to as a solid-state gain medium of
extremely favorable thermal properties, as strikingly proved by the
commercial availability of multikilowatt average power, single trans-
verse mode, near-diffraction-limited lasers. However, this advanta-
geous feature is inherently related to the possibility of distributing
thermal loads over long stretches of fiber, resulting in negligible ther-
mo-optical aberrations. Conversely, when the fiber must be shortened
to avoid NLE, the thermal load increases at a given pump and output
power. In such a situation, a very low core NA fiber may experience
thermally induced refractive index changes that are of comparable
magnitude with respect to the “cold” fiber core–cladding index step.