The detailed flame structure and velocity-forced flame response of lean-premixed, swirl-stabilized flames were investigated experimentally in a multi-nozzle can combustor. Measurements were obtained at atmospheric pressure in a laboratory-scale gas turbine combustor with five injectors, each containing a counter-clockwise swirler and a bluff centerbody, arranged in a" four-around-one" configuration. Velocity perturbations were introduced to the inlet fuel-air mixture at a fixed frequency and amplitude using a siren device. The velocity fluctuations were measured using the two-microphone method. CH* chemiluminescence was used as both a measure of heat release rate and as an indicator of the location of the flame. During the course of the study, a tomographic reconstruction technique was developed where line-of-sight chemiluminescence images from multiple angles around the combustor were used to reconstruct the three-dimensional chemiluminescence distribution of the multi-nozzle flame. Two-dimensional slices of the three-dimensional chemiluminescence image were studied in detail as they clearly reveal the interior structure of the multi-nozzle flame. Time-averaged 3-D chemiluminescence images were used to characterize the detailed flame structure of a stable, unforced flame with an overall flame shape typical to most inlet conditions at which the multi-nozzle combustor is operated. Two-dimensional slices of the 3-D images reveal the complex structure of the multi-nozzle flame that results from the five individual flames interacting with adjacent flames and the combustor wall. A comparison to the flow field in a similar multi-nozzle system showed that the merging of adjacent swirling flows affects the shape and strength of the recirculation zones which in turn affects the flame structure. Flame-flame interaction affects the heat release rate distribution within the flame brush but has a minimal effect on flame stabilization. Flame-wall interaction affects both flame stabilization and the distribution of heat release rate in the flame brush. Digital photographs and line-of-sight chemiluminescence images are used to study the evolution of flame structure with varying equivalence ratio. At a critical equivalence ratio close to the lean blowoff limit, an abrupt transition in flame structure occurs where the flame length increases, the heat release distribution changes and the middle flame is no longer anchored to the centerbody or dump plate. Time-averaged 3-D image data obtained at an equivalence ratio close to the lean blowoff limit show significant changes in flame structure when compared to the flame at operating conditions where all five flames are anchored to their respective centerbodies. The global response of the multi-nozzle flame to velocity perturbations was quantified using the flame transfer function. The input to the flame transfer function was the velocity fluctuation measured using the two-microphone method and the output was the heat release rate fluctuation of the entire multi-nozzle flame. Velocity fluctuation amplitude was maintained at 5% of the inlet velocity in order to remain in the linear flame response regime. Global flame transfer function data obtained at different equivalence ratios showed a change in the nature of the transfer function gain which was related to the change in overall flame shape. The local flame response was quantified using flame transfer functions of different regions of flame. Local flame transfer function data showed that the relative contribution of different regions of the flame towards the overall heat release rate fluctuation varies with equivalence ratio. Additionally, the phase relationship between the heat release rate …