Simulated canals.
Simulated canals made of clear polyester resin (Viapal uP 004/64, Vianova Resins, Hamburg, Germany) with coloured canal walls were used to assess instrumentation. The degree of curvature was either 288 or 358. The diameter and the taper of all simulated canals were equivalent to an ISO standard size 15 root canal instrument. The 288 canals were 13 mm long, the straight part being 5.5 mm and the curved part 7.5 mm. The curvature was defined mathematically with a radius of 7.5 mm resulting in an angle of 288, according to the Schneider method (Schneider1971). The 358 canals were 13 mm long, the straight part being 5 mm and the curved part 8 mm. The radius of curvature was 6.5 mm.

Preparation of simulated canals.
The simulated canals were prepared with either K3 rotary nickel-titanium instruments or with stainless steel hand K-Flexofiles. The transparent blocks were covered with adhesive tape during the preparation phase. All instruments were used to enlarge one canal only. Prior to use, each instrument was coated with glycerine to act as a lubricant, and copious irrigation with water was performed repeatedly after the use of each instrument. All canals were enlarged by an operator experienced in both manual instrumentation with stainless steel K-Flexofiles and automated preparation with K3 files. Measurement of the canals were carried out by a second examiner who was unaware of the experimental groups. A randomly laid down sequence was used to avoid bias towards one of the two instrumentation techniques. Only six resin blocks (three with 288 curves and three with 358 curves) were instrumented at a time to minimize operator fatigue and familiarity. These six resin blocks were defined as a set. The order of use of the two instrument types within a set was rotated. All canals were enlarged to an apical size 35. The following instrumentation sequences were used with the different instruments:
Group A: K3 instruments were set into permanent rotation (250 r.p.m.) with a 18 :1 reduction handpiece (K3 handpiece, W & H, Buermoos, Austria), powered by a torque-limited electric motor (K3etcm, Kerr, Karlsruhe, Germany) using torque setting 3,which is as stated by the manufacturer to be equivalent to a torque limitation of 1.2 N cm. Instrumentation was completed in a crown-down manner according to the manufacturer’s instructions using a gentle in-and-out motion. Instruments were withdrawn when resistance was felt and changed for the next instrument. In a pilot study the following instrumentation sequence allowed preparation of the different canals without difficulties:

1. A 0.06 taper size 20 instrument was used to 7 mm.
2. A 0.04 taper size 30 instrument was used to 9 mm.
3. A 0.06 taper size 25 instrument was used to 9 mm.
4. A 0.04 taper size 25 instrument was used to11mm.
5. A 0.04 taper size 20 instrument was used to 13 mm, the full length of the canal.
6. A 0.04 taper size 25 instrument was used to 13 mm, the full length of the canal.
7. A 0.04 taper size 30 instrument was used to 13 mm, the full length of the canal.
8. A 0.04 taper size 35 instrument was used to 13 mm, the full length of the canal.

Once the instrument had negotiated to the end of the canal and had rotated freely, it was removed.
Group B: Hand instrumentation with the K-Flexofile stainless steel instruments with noncutting tips was performed using a reaming motion. Thus, the instruments were manipulated in a clockwise rotation of about 90- 1208 until it reached the full working distance. A stepback method of instrument manipulation was not used. All canals were sequentially prepared from size 15 up to 35 without prebending the instruments, which were used to13 mm, the full length of the canal.
In each of these two test groups, 24 canals with 288 and 24 canals with 358 curves were enlarged. Thus, a total of 96 canals were prepared.

Assessment of canal preparation and analysis of data.
The time for canal preparation was recorded, and the total active instrumentation, instrument changes with in the described instrumentation sequence and irrigation was included. Changes of working length were determined by subtracting the final length (measured to the nearest 0.5 mm) of each canal after preparation from the original length (13 mm). The preparation time and the loss of working length were statistically analysed using the Mann-Whitney U-test at a significance level of P < 0.05. The number of fractured and permanently deformed instruments during enlargement was also recorded. A chi-square test was used to determine whether there were significant differences between the two instruments concerning instrument failure and deformation of instruments.
The assessment of preparation shape was carried out with the computer program Image 1.41 (National Institutes of Health public domain program). Therefore, pre- and postinstrumentation canal shapes were taken in a standardized manner and magnified 40 times by means of a charged coupled device (CCD) camera (SSCM370CE, Sony Corporation, Tokyo, Japan) and stored in the computer (Macintosh Quadra 660AV, Apple Computer, Ismaning, Germany). Thereafter, a composite image was produced of the pre- and postinstrumentation images and superimposed. The amount of resin removed, e.g. the difference between the canal configuration before and after instrumentation, was determined both for the inner and the outer side of the curvature in 1- mm steps using the Image1.41program. The amount of resin substance removed in all canals was measured one-dimensionally with a precision of _0.01mm. The first measuring point was 1mm away from the apical ending of the canal, the last measuring point was 10 mm from the apical end, resulting in 10 measuring points at the outer and10 points at the inner side of the canal, and thus, in a total of 20 measuring points (Schafer et al. 1995, Tepel 1997), all measurements were made at right angles to the surface of the canal. The data were analysed by the Mann-Whitney U-test, because for some measuring points the data was not distributed normally according to the Kolmogorov-Smirnov test.
Furthermore, based on the superimposition of pre and postoperative images, assessments were made according to the presence of different types of canal aberrations, such as apical zip associated with elbow, ledge and perforation. These different types of canal aberrations were defined according to the detailed descriptions published recently (Thompson & Dummer 2000b).